JP3789126B2 - Image recording device - Google Patents

Description

  The present invention relates to an image recording apparatus for recording a gradation image by an electrophotographic recording method.

  An electrophotographic printer has been widely put into practical use as a typical image recording apparatus for recording a gradation image (halftone image) as a hard copy. In this type of printer, the surface of the photoconductor is scanned with a light source such as a solid light emitting element array or a laser to form an electrostatic latent image on the photoconductor, and an image is formed by developing it with charged toner.

  In such an electrophotographic printer, binary recording is used by controlling whether or not to form dots on recording paper. In general, the relationship between the exposure intensity on the photoconductor and the recording density is non-linear and the gamma value is high, and the recording density changes greatly with a slight change in the intensity of the irradiated light. It is difficult. It is known to use a process such as a dither method or an error diffusion method when recording a shaded image, that is, a gradation image, with such a binary recording printer. These methods use the human visual characteristic that sensitivity is low for high spatial frequencies. By controlling the ratio of recorded pixels in multiple pixel groups, intermediate density corresponding to the ratio is expressed. It is a method to make it.

  The dither method is a method of expressing gradation using an image matrix, and has an advantage that can be realized with a relatively simple configuration. The error diffusion method binarizes the input level of one pixel, determines the recording level, and distributes the error between the input level and the recording level to the processed image only, thereby storing the density of the recorded image, An image having continuous gradation is obtained.

  Furthermore, a multi-value error diffusion method has been proposed in which this error diffusion method is extended to multi-value recording. However, the dither method allows stable gradation expression, but basically, recording is performed with the same number of pixel groups as the number of gradations as one unit, and thus a repeating pattern with a long period that appears on the recorded image is conspicuous. On the other hand, in the error diffusion method, the period of the pattern appearing on the recorded image is small in principle, but it becomes a special striped pattern and the image quality is greatly disturbed. In order to solve the problems of the error diffusion method, a technique for making a special pattern inconspicuous by adding a periodic signal to an image signal is known.

  In addition, the dither signal, which is a periodic signal, is added to the image signal to reduce the occurrence of special patterns, and the type of the image signal is identified, and the addition of the dither signal is reduced for images that require a sharp image. Techniques for processing are also known.

  However, in any of the above two methods, particularly when a halftone image is expressed, resolution degradation caused by adding periodic signals is caused. There is also known a method of reducing texture noise due to generation of a special pattern by performing multi-value recording instead of binary recording in error diffusion recording. Specifically, a method of reducing texture noise by recording with a multi-value printer by gradation processing using an algorithm similar to error diffusion is known. However, this method cannot be applied to a recording method that is not suitable for multi-value recording, such as an electrophotographic method. In other words, in electrophotographic recording, recording at a low density level is easily affected by temperature and humidity, and a stable density level cannot be maintained against fluctuations such as the environment and changes over time, resulting in an image with large density fluctuations. There was a problem such as.

  In order to improve image quality at low density, a technique is known in which binarization processing of lower bits is switched between binarization using a threshold value based on a random number and binarization based on a dither matrix depending on the value of the upper 1 bit of an input image. However, in the low density region, randomization is caused by binarization by random numbers, and dither patterns are conspicuous at medium density and above, and unnatural boundaries between low density and medium density are noticeable.

  A technique for improving halftone reproducibility by generating a checkered pattern by performing multi-level control for each dot is known, but extremely fine pulse width control is necessary to obtain sufficient gradation. It becomes necessary, and if there is a change in the environment, it becomes difficult to reproduce a stable gradation.

  Gives priority to each pixel in the image block to form a pixel, and gradation processing so that the pixel with the next priority starts growing even if the pixel with one priority is still growing continuously There is known a technique for making it strong against crushing, but this method is processing for each block, so resolution degradation is likely to occur, and sufficient gradation (for example, 64 gradations) or more per block is obtained. However, there is a problem that the recording system is required to have difficult performance such as sufficient stability of the recording system.

  Furthermore, recently, a method called a pulse width modulation method has been proposed in which a pulse width of a recording control signal supplied to a recording unit is modulated to express a gradation. In this method, one pixel is divided into a plurality of areas, and gradation is expressed by the number of areas used for recording, that is, the ratio of recording areas within one pixel. With this method, the spatial frequency of the pattern is sufficiently high, and the shading pattern is hardly noticeable.

On the other hand, however, there is a problem that density fluctuation and roughness are large in the low density portion. That is, in general, the spot of light irradiated on the photosensitive member is a spot having a certain size due to the size of the light source itself or the blur of the optical system. For this reason, as the pulse width of the recording control signal is narrowed, the potential distribution of the electrostatic latent image on the photosensitive member becomes a more distorted shape and deviates from binary recording. Therefore, the development density changes with a slight change in the environment, and the roughness and density fluctuation increase.
JP-A-4-207860 Japanese Patent Laid-Open No. 4-211569 JP-A-6-205218

  As described above, the conventional techniques have the following problems that cannot be solved. It is difficult to realize an ideal image recording apparatus capable of expressing one pixel with multiple values. In practice, unstable image formation is likely to occur by multi-valued recording within a pixel. For this reason, the multilevel error diffusion method is more advantageous in terms of smoothness of gradation than the binary error diffusion method, but image formation tends to be unstable.

  For example, when the recording system is inconvenient for forming a very small amount of image, the image formation in the highlight area becomes unstable, and an image with a lot of noise tends to be formed. Taking electrophotographic recording in which optical writing is performed by a semiconductor laser as an example of the recording system, it is known that the optical output of the semiconductor laser fluctuates as the environmental temperature varies. For this reason, image formation tends to be unstable, particularly in a low-density region, resulting in image quality with large density fluctuations. 43A is a conceptual diagram showing the relationship of the surface potential on the photoconductor, and FIG. 43B is a drive pulse width of the laser beam. As shown in FIG. 43, when a pulse width smaller than a predetermined pulse width is used, the surface potential on the photosensitive member becomes a level near the threshold value, so that image formation becomes unstable. That is, in the pulse width modulation, an intended density cannot be obtained, or pixel formation is unstable and the image quality is noisy.

  In addition, the sensitivity of the photoconductor varies due to manufacturing and changes over time, and pixel formation is unstable in multi-value recording, which is often recorded with a light amount at an intermediate level rather than binary recording. become.

  Furthermore, there is a problem that pixel formation is disturbed due to factors such as mechanical vibration due to driving of the photosensitive member, driving of the developing roller, and conveyance of the recording paper, and this point is also multivalued compared to binary recording. Large instability.

  The above problems are not limited to electrophotographic recording, and similar problems occur in other recording systems. Also, techniques such as the dither method and error diffusion method that express gradation images in recording systems that are not suitable for multi-value recording, such as conventional electrophotographic methods, have the problem that the resolution deteriorates and texture noise due to special patterns occurs. In addition, the pulse width modulation method solves the problem of reduced resolution and noise because the gradation changes in smaller units of the pixel size, but the problem is that the density change and roughness are particularly noticeable in the low density region. there were.

  The present invention has been made in consideration of the above problems, and provides an image recording apparatus capable of stably recording a good gradation image even for a general image recording apparatus that records an image with multiple values. With the goal.

  The present invention relates to an image recording apparatus capable of recording a good gradation image having high definition and low noise, and having no density change or roughness in a low density region, while using a recording system not suitable for multi-value recording. The purpose is to provide.

  In order to solve the above-mentioned problems, the present invention has taken the following measures.

Determining means for determining whether the input multi-value color image signal is a part of a line image or a pattern capable of improving sharpness ; and the multi-value pixel density value of the target pixel of the multi- value color image signal , The multi-value color image signal improves a part of the line image or sharpness by the signal conversion means for converting into a plurality of multi-value recording control signals defining the recording width and recording position of each color of the pixel of interest and the discrimination means. If it is determined that the pattern is a pattern that can be performed, the recording positions of the respective colors defined by the plurality of multi-value recording control signals are independently changed, and the multi-value color image signal has a part of the line image or sharpness. If it is determined not to be a pattern that can improve, and changing means does not change the recording positions of all colors that are defined by a plurality of said multi-level recording control signal, the multilevel recording system that has been changed by said changing means Recording means for multi-value recording an image based on the signal, the recording density estimating means for outputting an estimated value recording density estimate the recording density in the recording means from the multilevel recording control signal is changed by said changing means And an error diffusion means for diffusing an error between the multi-value pixel density value and the multi-value recording density estimation value into the multi-value color image signal.

  As described above, in the present invention, the recording density of an image point that is considered to be actually recorded is estimated from the recording control signal, and an error between the recording density estimation value and the pixel density value of the image signal is calculated to calculate the input image signal. To the surrounding pixels. As a result, the density of the low frequency component of the image is expressed by a plurality of pixels, but the high frequency component is expressed in units of one pixel, and high-definition recording is performed by recording position control at the edge portion.

  Depending on the type of the input image signal, for example, for a pattern in which the type of image can improve part of the line image or sharpness, the sharpness can be improved by changing the recording position defined by the recording control signal. Will improve.

  Furthermore, when the pixel density value of the target pixel is converted into a recording control signal, it is possible to control the image dot of the recorded image to have an appropriate size by referring to the recording control signal of the peripheral pixels of the target pixel. Thus, it is possible to record an image with low granularity in a state where density stability is enhanced.

  According to the present invention, the pixel density value of the target pixel of the input image signal is converted into a recording control signal that defines the recording width and recording position of the target pixel, and then supplied to the recording unit to record the image. , By estimating the actual recording density from this recording control signal and diffusing the error between the recording density estimated value and the pixel density value into the input image signal, while using a recording system that is not suitable for multi-value recording, It is possible to record a good gradation image with high definition and low noise and free from density change or roughness in a low density region.

  Depending on the type of the input image signal, for example, for a pattern in which the type of image can improve part of the line image or sharpness, the sharpness can be improved by changing the recording position defined by the recording control signal. Can be improved.

  Furthermore, when the pixel density value of the target pixel is converted into a recording control signal, by referring to the recording control signal of the peripheral pixels of the target pixel, it is possible to control the image point of the recorded image to be an appropriate size. It is possible to record an image with low granularity in a state where density stability is enhanced.

  Embodiments of the present invention will be described with reference to the drawings.

  (First Embodiment) FIG. 1 is a block diagram of an image recording apparatus according to a first embodiment of the present invention. In FIG. 1, an image input unit 1 (for example, a scanner or an image memory) outputs an image signal of a halftone image, that is, an image signal in which each pixel has a multivalued pixel density value. This image signal is input to the adder 2. The adder 2 outputs the corrected pixel density value by adding the corrected density value to the input image density value.

  The first signal conversion unit 3 uses the recording unit 5 to be described later to record the pixel density value output from the adder 2 in a first recording control signal that can form a pixel, specifically, the recording width of the recording pixel, It is converted into a first recording control signal that defines a recording amount such as a recording position. Although a detailed method for determining the first recording control signal will be described later, the first recording control signal is a provisional signal for recording an image in accordance with the input pixel density value. Input to the converter 4.

  The second signal conversion unit 4 refers to the recording control signal of the peripheral pixel adjacent to the target pixel from the first recording control signal that defines the recording amount, so that the recording unit 5 can form a stable pixel. In addition, a corrected recording control signal (second recording control signal) is output. A method for determining the second recording control signal will be described later. The second recording control signal is input to the recording unit 5 and is also input to the control signal buffer 6 and the subtracter 7.

  The recording unit 5 uses, for example, an electrophotographic recording unit that exposes and scans a laser beam onto a photosensitive member to form an electrostatic latent image and develops it with toner to obtain a recorded image. The present invention can also be applied to, for example, thermal recording using a thermal head and ink jet recording in which liquid ink is ejected, but in the present embodiment, description will be made below by electrophotographic recording means.

  The control signal buffer 6 is a buffer for temporarily holding the second recording control signal, and is configured by a shift register, a line memory, or the like, and receives a second recording control signal for a pixel adjacent to the target pixel. The recording control signal of the adjacent pixel corresponding to the target pixel is output to the second signal conversion unit 4 in a timely manner by a timing control signal (not shown).

  Of the second recording control signal, a signal corresponding to the recording width is input to the subtracter 7 and is subtracted from the corrected density value output from the adder 2, thereby generating an error. A signal is formed. At this time, the value input to the subtractor 7 is standardized to a data width equivalent to the pixel density value input to the first signal converter 3 by a standardization circuit (not shown). This error signal is input to the error diffusion unit 8.

  The error diffusing unit 8 includes an error buffer 8a that temporarily stores an error signal, and a weighting unit 8b that multiplies the error signal read from the error buffer 8a by a predetermined weighting coefficient, and accumulates the result. By supplying to the adder 2, error diffusion is performed on the image signal from the image input unit 1.

  In the following, the functions of the image input unit 1 and the recording unit 5 among the above-described units in FIG. 1 will be described in detail. For example, the image input unit 1 outputs a pixel density value of a bit map of a halftone image stored in an image memory as an image signal for raster scanning. Here, the bitmap is obtained by finely dividing an image into rectangular cells corresponding to pixels, and pixel density values are stored in each cell. The pixel density value is expressed as, for example, a digital value of 8 bits per pixel (that is, 256 gradations). The pixel density value of this bitmap is written into the image memory by the shake hand method from the host computer, for example. Further, the recording apparatus may have a CPU, and the information expressed in the page description language to be written may be developed into a bitmap from the host computer.

  FIG. 2 is a diagram illustrating a schematic configuration of the recording unit 5. The recording unit 5 of the present embodiment is an example of a recording unit that employs an electrophotographic system. In the recording unit 5, a charger 41, an exposure unit 22, a developing unit 42, a transfer roller 43, a charge removal lamp 46, and a cleaner 47 are disposed along the periphery of the photosensitive drum 24. The photosensitive drum 24 rotates at a constant speed in the direction of the arrow (clockwise) in the drawing based on a driving force transmitted from a motor (not shown) or the like. Since the configuration and operation of each part can refer to a conventional electrophotographic printer, the operation will be briefly described here.

  Prior to image formation on the recording paper 45, the photosensitive drum 24 starts to rotate. First, the surface of the photosensitive drum 24 is uniformly charged by the charger 41. Next, the light emission of the laser diode is controlled by the exposure control unit 22 based on the recording control signal generated based on the image signal, and the laser diode performs exposure scanning on the charged photosensitive drum 24 via the optical imaging system. .

  As a result of this exposure, the surface potential of the photoconductive drum 24 changes in the portion irradiated with light, and the surface potential varies depending on the portion not irradiated with light or according to the intensity of light irradiation. And an electrostatic latent image corresponding to the image signal is formed.

  FIG. 3 shows an example of the relationship between the exposure amount of the photosensitive drum 24 and the surface potential. As shown in FIG. 3, a potential pattern (that is, an electrostatic latent image) having a distribution corresponding to the recording control signal is formed on the photosensitive drum 24.

  In the developing device 42, the toner stored in the developing device 42 is attached to the photosensitive drum 24 corresponding to the electrostatic latent image as the developing roller 42a rotates, and a visible image is obtained.

  An example of the relationship between the surface potential of the photosensitive drum 24 and the toner adhesion amount is shown in FIG. Further, the recording paper 45 is transported between the transfer roller 43 and the photosensitive drum 24 with the timing controlled by a transport mechanism (not shown), and the toner developed by the developing device 42 is transferred onto the recording paper 45. Further, the recording paper 45 is conveyed to a fixing device 44 composed of a heat roller and the like, heated and pressurized to fix toner on the recording paper 45, and an image pattern corresponding to the image signal supplied to the exposure unit 22 is formed. Obtained on the recording paper 45.

  On the other hand, the photosensitive drum 24 is neutralized by a neutralizing lamp 46, and residual toner remaining without being transferred to the recording paper 45 side is removed by a cleaner 47. FIG. 5 shows the relationship between the exposure amount on the photosensitive drum 24 and the toner adhesion amount. The relationship between the two varies depending on factors such as the amount of charge and temperature, and changes from a solid line to a broken line, for example. In the region (2) where the exposure amount is small, the variation in the toner amount due to these variations is zero or small, but in the intermediate region (T1 to T2), the toner adhesion amount is easily affected by these variations. Therefore, if the medium exposure range (T1 to T2) is wide, the reproducibility of the recording density becomes unstable. Specifically, if such a variation occurs between pages, it results in a density variation, and if it occurs within a page, it causes roughness, density unevenness, and the like. The roughness of the former deteriorates the density reproducibility (color reproducibility in the case of color) and the density unevenness of the latter causes a reduction in image quality. Therefore, it is desirable to suppress these fluctuations to be small.

  FIG. 6 is a diagram showing a schematic configuration of the exposure unit 22. The structure of each element can be referred to a conventional laser exposure system. The pulse width converter 51 modulates the driving time of the laser diode in accordance with the recording control signal that is the output of the signal converter. A gradation image can be recorded by pulse width modulation. The light beam from the laser diode 52 driven in accordance with the modulation signal is deflected by a polygon mirror 54 via a condenser lens 53 including a collimator lens, a slit, and a cylinder lens, and f− as an optical imaging system 55. Exposure scanning is performed on the photosensitive member 24 through the θ lens.

  Next, the operation of each unit will be described according to the processing procedure. The pixel density value for the target pixel input to the first signal conversion unit 3 is converted into a first recording control signal suitable for recording in the recording unit 5 in accordance with the magnitude of the density value. For example, a case where the recording unit 5 records five values from 0 to 4 as shown in FIG. 7 will be described. First, 256-level input pixel density values are converted into one of five levels. Output recording control signal. This conversion can be realized by an equal interval quantization operation, and the first signal conversion unit 3 as shown in FIG. 8 can be a lookup table using a ROM or the like. FIG. 8 is a block diagram showing an example thereof, and FIG. 9 is an example of a processing table.

  The second signal conversion unit 4 refers to an adjacent pixel control signal that is a recording control signal of at least one of neighboring pixels adjacent to the target pixel, and inputs the input first recording control signal to the target pixel. And a second recording control signal capable of stable recording. FIG. 10 is a block diagram illustrating an example of the second signal conversion unit 4 and the control signal buffer 6. The second signal conversion unit 4 can be realized by a look-up table using a ROM or the like similarly to the first signal conversion unit 3.

  For example, when the adjacent pixel to be referred to is the left adjacent pixel of the target pixel, that is, the previous pixel, the second recording control signal to be output with reference to the table shown in FIG. 11 is determined. That is, in FIG. 11, the first recording control signal (current pixel) for the target pixel is “1” or “2”, and the second recording control signal (previous pixel) corresponding to the previous pixel is “0” to “0”. In the case of “3” (in the box in the figure), correction processing is performed. This is a case where a recording pattern such as “1” or “2” shown in FIG. 7 is selected. In this example, a minute value such as “1” or “2” is used alone as a recording control signal. Since the recording stability when selected is insufficient, correction processing is performed so as to avoid recording using this value. However, even if the current pixel is a recording control signal of “1” or “2”, if the recording control signal of the previous pixel is “4”, the recording pattern of adjacent pixels is continuously as shown in FIG. Since pixels are formed, stable recording can be performed. Therefore, in this case, the correction process is not performed, and the first recording control signal for the current pixel is used as it is as the second recording control signal.

  For the conversion from the first recording control signal to the second recording control signal, the value of the recording control signal closest to the value of the first recording control signal and capable of stable recording is selected. The reason for selecting the value closest to the value of the first recording control signal is to realize faithful gradation reproduction while minimizing the error from the recording density for the pixel of interest. When there is a variation in instability, a recording control signal that allows stable pixel formation at any time is selected in consideration of the allowable range of variation.

  As a method of this conversion, the size of a minute pixel on which unstable recording is performed can be obtained in advance by experiment. When the minimum recording control amount for performing stable recording is Dstable and the pixel density value corresponding to the first recording control signal for the target pixel is D1, correction processing is performed with reference to D1. In this correction process, for example, the first recording control signal is binarized with a threshold value of Dstable / 2 to obtain a second recording control signal.

  At this time, even when a value different from the initial first recording control signal is selected, the second recording control signal, which is the final output, is supplied to the difference from the input pixel density value via the error diffusion unit 8. Is diffused to other pixels as an error, so that the macro density of the input pixel is preserved.

  Further, a table as shown in FIG. 13 or FIG. 14 may be found and applied from an experiment for obtaining recording stability. The table in FIG. 13 is an example in which, when the recording control signal for the previous pixel is “3”, recording of a minute pixel is stably performed with a recording control signal of about “2” for the target pixel. In this case, the previous pixel and the current pixel do not give a completely continuous pattern, but in the pixel forming operation in the recording system, for example, considering that the exposure pattern by the laser beam is blurred to some extent, This is an example in which the continuity of the recording pattern is obtained when the above is comprehensively considered. The table in FIG. 14 is an example when applied to a recording system with relatively large instability. That is, in this example, it is assumed that unstable pixel formation occurs when a recording pattern of about “3” shown in FIG. 7 is selected alone as the recording control signal for the current pixel.

  As described above, various conversion methods can be selected from the first recording control signal to the second recording control signal. However, the gist of the present invention is that the recording control signal (only the unstable control pixel is formed) In this case, the second recording control signal) is to be removed.

  In this way, the unstable pixel is not selected at all, but the formation of the minute pixel is selected in accordance with the pixel formation state by referring to the pixel formation state with the peripheral pixels. Smooth gradation characteristics can be maintained. In addition, since there are control levels that cannot be expressed in the highlight area, the gradation is poor, but a stable and low noise image can be obtained, which has the advantage of improving the overall image quality.

  In addition, when the reference table shown in FIG. 11 or FIG. 14 is applied, the binary signal indicating whether the previous pixel is “4” or other than the previous pixel recording control signal information. That is, since it is sufficient to use 1-bit information as the data amount, the control signal buffer 6 can have a simple configuration. That is, when the second recording control signal for the previous pixel of the target pixel is temporarily held as an adjacent pixel, it can be realized with a simple configuration such as a 1-bit latch (see FIG. 10).

  Next, the function of the error diffusion unit 8 will be described in detail. The error signal from the subtracter 7 is stored in an error buffer 8a comprising a line memory. The error signals stored in the error buffer 8a are sequentially read out and input to the weighting unit 8b, and are adjacent to the pixel around the pixel of interest X shown in FIG. 15 (that is, three pixels A, B, and C one line before the left). Obtained by multiplying the error signal of the previous pixel D) by the weighting coefficients (Wa, Wb, Wc, Wd), respectively, and integrating them. For example, Wa = 1/16, Wb = 5/16, Wc = 3/16, and Wd = 7/16 are selected as the weighting coefficients. As a simple example, the numerator may be set to a power of 2 such as Wa = 2/16, Wb = 4/16, Wc = 2/16, and Wd = 8/16. In this case, since the weighting unit 8b can be configured only by bit shift and addition, the circuit is simplified. In this way, the weighted and integrated signal is added to the image signal input by the adder 2, whereby the quantization error during recording is diffused.

  In the above description, the previous pixel with respect to the target pixel is used as the peripheral pixel to be referred to. However, in addition to the previous pixel, a pixel on the previous line (a pixel located on the previous line with respect to the target pixel) is included as the peripheral pixel to be referred to. Fine control is possible.

  FIG. 16 is a block diagram showing a specific example of the second signal converter 4 and the control signal buffer 6. The second signal conversion unit 4 is configured by the ROM 4, and the control signal buffer 6 is configured by a line memory 6a corresponding to the number of pixels for one line and a latch 6b for one pixel. That is, of the second recording control signal, a value indicating the recording width is input to the latch 6b as a representative recording amount. If the i-th pixel is the target pixel, the second recording control signal for the (i−1) -th pixel is stored in the latch 6b. The output value of the latch 6b and the recording control signal for the pixel of the previous line stored at the i-th address of the line memory 6a are simultaneously read out and output as the first signal converter. At the same time as one recording control signal, it is input to the ROM 4 which is the second signal converter 4. For these data, the address of the ROM 4 is selected, and a second recording control signal corresponding to this address is output. The output of the latch 6b is stored at the (i-1) th address of the line memory 6a, and the process proceeds by repeating the same operation. Here, a configuration using a FIFO instead of the line memory 6a may be employed. In this case, there is an advantage that a control circuit for address selection of the line memory becomes unnecessary.

  FIG. 17 is a reference table showing an example of how to determine the second recording control signal with respect to the first recording control signal. In the example shown in FIG. 17, when the previous pixel is “0” to “3” and the previous line is “0”, an unstable pixel is formed, and thus correction processing is performed. When the previous pixel is “4” or the previous line is “1” to “4”, the first recording control signal is output as it is as the second recording control signal, assuming that stable pixel formation is performed. . These look-up tables are determined in consideration of the continuity or the degree of connection between the previous pixel, the previous line pixel, and the target pixel, and can be determined accurately through experiments performed in advance.

  In the above description, an example in which the first recording control signal is output by the first signal converting unit and the second recording control signal is obtained by the second signal converting unit based on the first recording control signal has been described. Can also be performed together in one process as a series of operations. That is, a circuit configuration such as a ROM that is set to output the second recording control signal corresponding to the pixel input signal with reference to the adjacent pixel control signal may be used. Even in this case, the ROM table can be determined in accordance with the above description.

  In the above description, the recording of five values from “0” to “4” has been described. However, the present invention is not limited to the number of multi-values of recording, and the smoothness increases as the number of multi-values increases. Tonal gradation can be realized. Conventionally, even in a recording apparatus in which a multi-valued number cannot be set large in order to avoid unstable pixel formation, by applying the present invention, a stable gradation with rich gradation using a larger multi-valued number is obtained. There is a merit that an image can be recorded.

  (Second Embodiment) FIG. 18 is a block diagram of an image recording apparatus according to a second embodiment. The same parts as those in the first embodiment are denoted by the same reference numerals, and detailed description thereof is omitted. In the second embodiment, the control signal buffer 6 of the first embodiment is not provided and a recording density estimation unit 9 is provided. Note that the signal conversion unit includes only the signal conversion unit 3 corresponding to the first signal conversion unit 3 of the first embodiment.

  The image input unit 1 and the adder 2 have the same functions as in the first embodiment. The signal conversion unit 3 is a part corresponding to the first signal conversion unit 3 of the first embodiment. The signal conversion unit 3 quantizes the pixel density value output from the adder 2 and determines the recording width, recording position, and the like of the pixel. It is converted into a recording control signal that defines the recording amount. Although a detailed method for determining the recording control signal will be described later, it is generally a signal for recording the input pixel density value, and is input to the recording unit 5 and the recording density estimation unit 9.

  Since the function of the recording unit 5 is the same as that of the first embodiment, the description thereof is omitted. The recording density estimation unit 9, which is a feature of the second embodiment, estimates the actual recording density in the recording unit 5 based on the recording control signal output from the signal conversion unit 3, and the recording density that is the estimation result Output the estimated value. This recording density estimated value is input to the subtractor 7, and the subtracting record 7 subtracts the recording density estimated value from the corrected density value output from the adder 2 to generate an error signal. This error signal is input to the error diffusion unit 8.

  The error diffusion unit 8 is the same as in the first embodiment. In the following, details of the function of each unit in FIG. FIG. 19 shows the configuration of the recording unit 5. In the first embodiment, a scan type recording unit is shown as the recording unit 5, but the recording unit 5 of the present embodiment is an array type recording unit. The recording unit 5 mechanically drives light from the solid-state light emitting element array 22 arranged in the main scanning direction driven by the drive circuit unit 21 in the sub-scanning direction (recording paper feeding direction) by the fiber type lens array 23. An image is formed on the rotating photosensitive drum 24. The drive circuit unit 21 converts the time-series recording control signal input from the signal conversion unit 3 into a signal for simultaneously driving the light emitting elements of the solid state light emitting element array 22. The solid state light emitting element array 22 is arranged in parallel with the rotation axis of the photosensitive drum 24 and close to the surface of the drum 24. The fiber lens array 23 is disposed between the solid light emitting element array 22 and the photosensitive drum 24 so as to form an image of the light emitting surface of the light emitting element on the photosensitive drum 24. As a result, light from the solid state light emitting element array 22 is irradiated onto the photosensitive drum 24 as exposure light, and the surface of the photosensitive drum 24 is exposed. As the solid-state light emitting element array 22, an LED print head in which LEDs are arranged in a row, or one using light emission of a phosphor can be used.

  FIG. 20 shows the configuration of the drive circuit unit 21 of the solid-state light emitting element array 22. The recording control signal from the signal conversion unit 3 is first written sequentially into the shift register 31 line by line. The shift register 31 is configured by cascading a plurality of 5-bit parallel shift register elements. The contents of the shift register 31 are output to the latch 32 in parallel. In this embodiment, each of the shift register 31 and the latch 32 includes a plurality of IC chips, and the plurality of IC chips are connected in cascade.

  Of the data of the recording control signal latched in the latch 32, 4-bit data representing the recording width (exposure pulse width) is compared with two sawtooth wave signals S3 and S4 having different polarities in the two comparators 33 and 34. Is done. One of the output signals of the comparators 33 and 34 is selected by the selector 37, and the signal selected by the selector 37 becomes a drive signal to the corresponding element of the solid-state light emitting element array 22. The selector 37 selects one of the outputs of the comparators 33 and 34 based on 1-bit data representing the recording position (exposure pulse start timing) among the data of the recording control signal latched by the latch 32. Is controlled.

  In this way, the light emitting element array 22 emits light for the time of the exposure pulse width defined by the recording width data of the recording control signal at the exposure pulse start timing defined by the recording position data of the recording control signal for each light emitting element. To be controlled.

  The signal conversion unit 3 will be described in detail. In the recording unit 5, the light emission time (exposure pulse width) of the solid state light emitting element array 22 is controlled in accordance with the sub-scanning (rotation of the photosensitive drum 24), so that the recording width in the sub-scanning direction within one pixel is controlled. By changing the recording density and changing the recording position by controlling the light emission start timing (exposure pulse start timing) of the solid state light emitting element array 22, it is possible to record a gradation image with a stable density. The signal conversion unit 3 uses the pixel density value of the image signal input from the adder 2 as a recording control signal that defines the recording width and recording position in the recording unit 5, that is, the exposure pulse of each pixel in the solid-state light emitting element array 22. The information is converted into width and exposure pulse start timing information.

  The function of the signal converter 3 will be described in more detail. The signal conversion unit 3 has the following two functions. The first function is a function of expressing the recording density of the number of levels such as 256 gradations by a recording control signal in which the number of levels is limited to about several levels in pixel units. The second function is a function that prevents the size of the connected image dots in the toner image formed on the photosensitive drum 24 from becoming too small.

  First, a mechanism for expressing a recording density of 256 gradations will be described. In the second embodiment, the recording density is controlled by the size of the latent image distribution on the photosensitive drum 24, that is, the pulse width (exposure pulse width) of the exposure light irradiated on the photosensitive drum 24. However, in general, the exposure pulse width for recording one pixel is usually on the order of a few microseconds or less, and if the exposure pulse width is controlled in units finer than this, it is necessary to control at a higher speed by the number of divisions. As a result, the circuit cost becomes high and the operation becomes unstable, so that the number of divisions cannot usually be made too large. For example, if the exposure pulse width is divided into 15, the exposure pulse width can be controlled only in 16 ways per pixel, and a recording density of 256 gradations cannot be expressed. Further, since the relationship between the exposure pulse width and the recording density is actually non-linear, even if the exposure pulse width is divided into 256, the recording density cannot be uniformly controlled in 1/256 level steps.

  As described above, the error diffusion method is used in the present embodiment as a method for expressing a sufficient gradation when the number of control levels (in this case, the number of divisions of the exposure pulse width) is limited. As is well known, the error diffusion method is a method that utilizes the human visual characteristic that gradation resolution is low at high resolution. By distributing the quantization error at the target pixel to adjacent pixels, a set of multiple pixels is obtained. The desired recording density is reproduced in units of. This error diffusion is performed based on the output of the error diffusion unit 8 in the adder 2.

  FIG. 21 shows a pixel arrangement of an image recorded on the recording paper 45 by the recording unit 5. In FIG. 21, the horizontal direction represents the main scanning direction (the arrangement direction of the solid-state light emitting element array 22), and the vertical direction represents the sub-scanning direction (the rotation direction of the photosensitive drum 24 = the feeding direction of the recording paper 45). The numbers arranged in the main scanning direction and the sub scanning direction respectively represent pixel positions (X) and (Y) in the main scanning direction and the sub scanning direction, and a plurality of pixels in the sub scanning direction (four in the example in the figure). The numbers shown in the divided areas represent the position of the area.

  In this case, the light emission start timings for the odd-numbered (1, 3,...) Main scanning lines and the even-numbered (2, 4,. In the odd-numbered main scanning line, the recording width is increased using the upper end of the pixel in the drawing as a reference (hereinafter referred to as “pre-reference”), and in the even-numbered main scanning line, the lower end of the pixel in the drawing is used as the reference (hereinafter referred to as “below”). The gradation image is recorded by increasing the recording width as “post-reference”.

  In this way, it is possible to express multiple gradations by controlling the recording width within one pixel, and by concentrating the image points in the low density region by controlling the recording position, density stabilization and variation noise can be reduced. Reduction can be achieved.

  FIG. 22 shows detailed configurations of the signal conversion unit 3 and the recording density estimation unit 9 of FIG. The signal conversion unit 3 can be realized by a ROM as shown in FIG. For example, a table as shown in FIG. 23 is stored in the ROM used for the signal conversion unit 3. In FIG. 23, “pixel density value” is a pixel density value of each pixel of an image signal input to the signal conversion unit 3 from the image input unit 1 via the adder 2 and has a value of 256 levels. “Pixel position” represents X and Y in FIG. 21. In this case, only the pixel position Y in the sub-scanning direction is used. “Output” is a recording control signal output from the signal conversion unit 3. In the example shown in FIG. 23, for example, a 3-bit recording width representing a recording width of 0 to 4 corresponding to the pixel array shown in FIG. Information (indicated by “width” in FIG. 23) and 2-bit recording position information (indicated by “position” in FIG. 23) representing recording positions 1 to 4. That is, the recording width information is information indicating how many of the four areas in one pixel are used for recording in FIG. 21, and the recording position information is information indicating from which position of these four areas recording is started. is there.

  The recording density estimation unit 9 will be described. The recording density estimation unit 9 estimates the recording density in the recording unit 5 based on the recording control signal output from the signal conversion unit 3 as described above, and outputs a recording density estimation value. In general, in a recording system in which a recording image slightly blurs like an electrophotographic system, the recording density of each pixel is affected by adjacent pixels. Therefore, the recording density estimation unit 9 determines the estimated recording density from the data of the four adjacent pixels of the recording control signal output from the ROM of the signal conversion unit 3, as shown in FIG.

  That is, the recording control signal is input to the line memory 56 and the one-pixel delay latch 57, and the data of the two pixels on the line one line before the main scanning line to which the current pixel (target pixel) belongs is extracted from the line memory 56. The recording density determination unit 58 calculates and determines the recording density of the current pixel from the data D4, D2, the data D3 of the previous pixel taken out from the latch 57, and the data D4 of the current pixel.

  The most accurate method for determining the recording density is to actually perform a recording experiment using all combinations of the four-pixel data D1 to D4, measure the recording density based on the experimental result, and store the recording density measurement result in the ROM. It is a method to store in. However, in this example, since the recording control signal is expressed by data of 5 bits per pixel, if data of 4 pixels is used for determining the recording density, the recording density data stored in the ROM becomes a combination of 20 bits. The scale of ROM becomes considerably large. In this embodiment, since one pixel is divided into four areas as shown in FIG. 21, a 16-bit combination is possible depending on the device, but the ROM size is still large.

  According to the results of actual recording experiments, even if the recording position (image point position) of each pixel in the recorded image changes slightly, the recording density does not change much. Therefore, the recording control signal is separated into recording width information and recording position information, and for the recording width information, the sum of the recording width information of the data D1 to D4 of 4 pixels (corresponding to the density sum of 4 pixels) is obtained. For the recording position information, only 1 bit per pixel, that is, only information for distinguishing the front reference and the rear reference is extracted, and the sum of the recording width information of 4 pixels is corrected as 4 bits of information by 4 pixels. Accurate estimation of recording density is possible.

  FIG. 24 is a diagram showing a configuration of the recording density estimation unit 9 simplified by such a method. The 4-pixel data D1 to D4 (see FIG. 22) of the recording control signal are input to a 5-bit input ROM (which may be a logic circuit) 61 to 64, and before the 3-bit recording width information and the recording position information. The 1-bit data that distinguishes the reference and the post-reference is separated and extracted. The 1-bit data of the recording position information to be output from the ROMs 61 to 64 is obtained from the “width” and “position” information in FIG. That is, 1-bit data of recording position information is 1 regardless of “width” if “position” is 1, 1 if “position” is 2 and “width” is 1 or 2, and “position” is 3 0 if the “width” is 1, 0 if the sum of the “position” and “width” is 5, and if the “position” is 1 and the sum of the “position” and “width” is 5. It is set as “0” or 1.

  The 4-bit data obtained by adding the 3-bit data of the recording width information by the adder 65 and the 1-bit data of the 4-pixel recording position information are input to the 8-bit input ROM 66, and the former data is corrected by the latter data. The obtained 8-bit recording density data is obtained. The ROM 66 can be created by storing 16 levels of combinations of 16 pre-reference and post-reference with 4 pixels, that is, recording density measured by performing 256 recording experiments. it can. Actually, since the recording density is the same for the symmetric combination, it is not necessary to perform the recording experiment for all the combinations, and it is sufficient to perform the 1/4 recording experiment of 1/4.

  This method can be similarly applied even when the number of divisions of one pixel is increased from four. For example, in the case of dividing into eight, the reference of the recording position is the same, and the recording width information is only increased by one bit. The ROM for determining the recording density may be a 9-bit input ROM, and even if one pixel is divided into nine, a 10-bit input ROM may be used. Even when the recording density estimation accuracy is increased, the hardware scale There is almost no increase.

  The density estimation value obtained by the recording density estimation unit 9 as described above is obtained by the subtracter 7 to obtain an error from the corrected pixel density value output from the adder 2, and this error signal is used as the pixel of interest. It is diffused to neighboring pixels, and the sum is added to the pixel density value of the next pixel by the adder 2. By such error diffusion processing, even if an error occurs between a pixel density value and a recording density value in a certain pixel, the error is transferred to an adjacent pixel, so that the density in a macro region is reproduced. .

  Generally, an error occurs between the estimated recording density and the actual recorded density. This is due in part to the fact that the number of control levels of the recording width (exposure pulse width) specified by the recording control signal is smaller than the required number of gradations. This is because a value close to the image density value is not selected. As will be described later, since the signal conversion unit 3 outputs a recording control signal in which the size of the image dot becomes a certain size, the recording width that reproduces the recording density closest to the micro level is not necessarily obtained. It is because it does not choose. However, since the error between the estimated recording density and the actual recording density is transferred to the next pixel by error diffusion, the recording density in the macro area can be correctly reproduced.

  As described above, in the present embodiment, fine pattern information, that is, the recording width and the recording position are determined for each pixel according to the pixel darkness value of the input image signal, which is sufficient for abrupt changes in the image signal. Response is possible, and high-definition images can be expressed.

  On the other hand, the recording density is estimated by taking into account the influence of adjacent pixels and density changes depending on the position within one image point, so error density can be diffused by feeding back recording density information with high accuracy. Thus, an extremely accurate concentration can be expressed.

  FIG. 25 is a diagram showing another example of a table stored in the ROM constituting the signal conversion unit 3. 25 differs from FIG. 23 in that the recording width information and the recording position information are changed not only by the pixel position Y in the sub-scanning direction but also by the pixel position X in the main scanning direction. When the signal conversion unit 3 is realized by a ROM storing the table shown in FIG. 23, the recording width information and the recording position information are the same in both the odd-numbered columns where X is an odd number and the even-numbered columns where X is an even number. Since the fine patterns have the same arrangement, the recorded image becomes a line pattern continuous in the main scanning direction, that is, a horizontal line pattern when expressing a halftone of uniform density.

  In contrast, as shown in FIG. 25, the ROM constituting the signal conversion unit 3 stores a table in which the recording width information and the recording position information are different between the odd and even columns and the arrangement of the fine patterns of the recorded images is different. Then, the recorded image becomes a line pattern extending in the sub-scanning direction at a low density, and becomes a pattern in which cross patterns are connected at a high density. That is, as shown in FIG. 26 (a), the line pattern slightly extends in the sub-scanning direction at a low density, and the line pattern continuous in the sub-scanning direction as shown in FIG. 26 (b) at a medium density. That is, the pattern is based on vertical lines.

  In the so-called contact development system in which the photosensitive drum 24 and the developing roller 42a of the developing device 42 are in contact with each other as shown in FIG. 2, there is a difference in peripheral speed between the photosensitive drum 24 and the developing roller 42a. The toner supplied to the photosensitive drum 24 from 42a is easily affected by the developing roller 42a, and the fine pattern may be deformed or missing. However, as shown in FIG. 26B, in the pattern that is continuous in the sub-scanning direction, that is, the direction in which the developing roller and the photosensitive drum 24 are separated, the fine pattern is less deformed and missing, and recording with less roughness noise is performed. It becomes possible. When the density is medium or higher, a cross-shaped pattern is formed as shown in FIG. 26 (c), and in this case, the toner adheres stably.

  According to such a fine pattern, relatively stable gradation expression can be achieved even in a developing system in which the photosensitive drum 24 and the developing roller 42a are in contact with each other, such as contact development with a one-component non-magnetic toner. Further, there is a characteristic that it is strong against unevenness of a recording image point due to the eccentricity of the photosensitive drum 24, and the roughness noise is reduced. Although the recording density tends to increase suddenly when the pattern of FIG. 26B is shifted to the pattern of FIG. 26C, the image is displayed in the main scanning direction slightly before being completely connected in the sub-scanning direction. By increasing the points, it is possible to mitigate such a rapid increase in recording density.

  FIG. 27 shows examples of various combinations of fine patterns of recorded images when the table of FIG. 25 is used, and the vertical and horizontal axes represent the relationship between the dot widths of even and odd columns, respectively. According to the meridian diagram shown in P2 of FIG. 27, first, the recording width over three areas is recorded among the areas divided into four in the sub-scanning direction in the odd columns, and then the recording width of the remaining one area in the even columns. Record the minutes. Next, recording is performed again for the recording width of one area in the odd-numbered columns, and further recording is performed for the recording width of one area in the even-numbered columns. By doing so, it is possible to reduce a sudden change in gradation.

  Also, as shown in FIGS. 26B to 26D, when the sub-scanning direction is completely continuous, the meridian of P1 in FIG. 27 is obtained. That is, recording is performed on the recording width dividing lines of the four areas in the odd columns, and then recording is performed up to the solid density of the recording width of the four areas in the even columns. Also, P3 in FIG. 27 is an example recorded in the table in FIG.

  By combining the recording levels in the odd and even columns in this way, various fine patterns can be combined, and it is possible to reduce roughness noise and stabilize the gradation characteristics according to the recording characteristics of pixel formation. It becomes possible.

  In this embodiment, the monochrome image recording apparatus has been described. However, a color image recording apparatus can also be applied, and a remarkable effect can be obtained. In that case, the recording unit records a color image using, for example, four colors of ink of yellow (Y), magenta (M), cyan (C), and black (K), but the recording control signal is for each of these colors. Generated independently. Further, when the position deviation of the color printer is as small as around one stroke or less, or less than one stroke, the “position” information in the table of FIGS. 23 and 25 is changed for each color of Y, M, C, and K. By doing so, stable color reproduction becomes possible. However, in an actual color printer recording system, a positional deviation of about 1.5 pixels often occurs, and even if the same recording position is controlled for each color, the actual color reproduction is an average recording image point. Since the position is determined by the position, the color reproduction color hardly fluctuates particularly greatly.

  (Third Embodiment) In the present embodiment, for a case where a halftone image and a character and line image are all processed and inputted as halftone data, the halftone image faithfully reproduces the halftone, and the character and line image are An example in which sharpness is preferentially reproduced will be described.

  As an image signal source (that is, an output of the image input unit 1), a printer that outputs a scanner output signal or halftone multi-value pixel data as it is and generates bitmap pixel data for a character or line drawing code signal. Assumes controller output. For example, when lines are recorded obliquely as shown in FIG. 28, an image signal source is one that generates halftone data of the area ratio at each image point. That is, in FIG. 28, Δ, ○, and □ are a small printing rate, intermediate printing rate, and solid printing rate, respectively.

  In this case, since the image is originally a line figure, simply increasing the recording width in the front reference and the rear reference in the odd and even lines, respectively, as in the second embodiment, expresses the line segment sharply. Then, it is not preferable. Therefore, in this embodiment, by examining the pattern (type) of the input image signal, it is determined whether it is a part of the line image or a pattern that can improve the sharpness, and the part of the line image or the sharpness is improved. If the pattern is possible, the pre-reference and the post-reference are determined so that the reproduction of the line image becomes more faithful, specifically, so that the line image is reproduced more sharply. If the pattern is not a part of the line image or cannot improve the sharpness, the pre-reference and the post-reference are determined by the odd lines and the even lines as in the second embodiment. That is, the recording position determined in the second embodiment is not changed.

  FIG. 29 is a block diagram showing a configuration of the image recording apparatus according to the present embodiment having such a function, and the same reference numerals are given to portions corresponding to those in FIG. In FIG. 29, the image discriminating unit 11 discriminating the type of the image signal from the image input unit 1 and the recording position information of the recording control signal output from the signal converting unit 3 according to the discrimination result of the image discriminating unit 11 are shown. A recording position changing unit 12 that changes the recording position by changing the position is provided.

  The image discriminating unit 11 discriminates the type of the image signal from the image input unit 1 from the image pattern, and determines whether to use the front reference or the rear reference based on the determination result. That is, when the image determination unit 11 determines that the image signal is a part of a line image or a pattern that can improve sharpness, the recording of the quantized recording control signal output from the signal conversion unit 3 is performed. The content of the position information is rewritten by the recording position changing unit 12. If it is not determined that the pattern is a part of a line image or a pattern that can improve sharpness, such rewriting of the recording position information is not performed.

  In addition, when the recording position information is rewritten by the recording position changing unit 12, the estimated recording density output from the recording density estimating unit 9 is corrected. That is, when the sharpness of the recorded image is emphasized and the density changes, the recording position information of the recording control signal input from the signal converting unit 3 to the recording density estimating unit 9 via the recording position changing unit 12 changes. Therefore, the address value input to the ROM shown in FIG. 22 changes accordingly, and the recording density estimation value is corrected, whereby the overall recording density is maintained by the error diffusion loop.

FIG. 30 shows a detailed configuration of the image determination unit 11. The image signals input from the image input unit 1 are delayed by the line memories 71, 72, 73, 74, and the same row of image data from the first line to the fifth line is taken out in parallel. The delayed image data of the third line is also input to the adder 2 in FIG. The image data of the first line to the fifth line are displayed at a density of 0 (x = 0) and a halftone (x = x) by multi-value / binary determination circuits 75, 76, 77, 78, 79 constituted by a ROM or a logic circuit. 1) is identified and converted into 2-bit data. The sharpness enhancement determination ROM 80 determines whether or not the sharpness can be enhanced by the 2-bit data of each line from the multi-value / binary determination circuits 75, 76, 77, 78, and 79, and the enhancement is possible. If so, the data of the front reference and the rear reference is output.

  FIG. 31 is an example of a table stored in the sharpness enhancement determination ROM 80, and shows the relationship between the input value p (p1 to p5) and the output value d. FIG. 32 shows how the recording position is changed in the upper and lower 3 pixels or 5 pixels. In FIG. 32, a broken line indicates a non-recording area, and a solid line indicates a recording area.

  In the example of FIG. 31, the recording position is changed only when the image signal is halftone (p3 = 1). This is because when the image signal is white or solid, the recording position is always the end of the pixel and cannot be changed. Further, when the halftone near binary data is white on the opposite side, it is an edge of a line segment, and the halftone is moved to the solid side. That is, when one of two adjacent image points is solid and the other is white (in the case of a and b in FIG. 31, that is, p2 = 2 and p4 = 0, or p4 = 2 and p2 = 0) In FIG. 32, (a1) is changed to (a2) and (b1) is changed to (b2) so that the recording position is changed to the solid image point side. In this way, the line is not blurred and the edge is sharpened, and the roughness noise is reduced in an oblique line that is nearly parallel to the main scanning direction.

  Also, if one of the adjacent dots is halftone and the other is white, and the dots adjacent to both sides are white, that is, a thin line segment is scanned across two dots or a rasterized image. In the case of (c and d in FIG. 31), the recording position is changed so that the line segment becomes thin. That is, in FIG. 32, the recording position is changed from (c1) to (c2) and (d1) to (d2). Other than these, the recording position is not changed. By doing so, it is possible to reduce the roughness noise of diagonal lines and to reproduce a sharper image.

  Note that the change of the recording position may be selected so that the reproduced image is sharper and further reduces the roughness noise, and is not limited to changing the recording position according to the table of FIG. In addition, the sharpness enhancement determination unit is not necessarily configured by a ROM, and may be configured by a logic circuit.

  In the present embodiment, the image type of the input image signal is directly determined from the input image signal to determine the effectiveness of changing the recording position. However, the recording control signal output from the signal conversion unit 3, that is, quantized. The image type of the input image signal may be determined from the received image signal, and the recording position may be changed based on the image type. Also in this case, the recording position changing process can be realized with the same configuration as that shown in FIG. However, in this case, instead of the input image signal from the image input unit 1, only the recording width information in the quantized recording control signal has to be used, so the storage capacity of the line memories 71 to 74 may be small. .

  Furthermore, in the present embodiment, the processing for obtaining a sharp feeling by changing the recording position according to the type of input image signal (arrangement of image patterns) has been described. If the processing of this embodiment is performed after emphasis, it becomes more effective. That is, the second and third embodiments are based on multi-level error diffusion. When such processing is performed, image data is generally converted into a blurred image. However, if high frequency emphasis is performed to compensate for this blur, a sharper image can be obtained. Further, in this case, high frequency emphasis is performed to identify whether or not the recording position needs to be changed, so that the edge can be detected more easily and sharpness can be effectively enhanced. .

  (Fourth Embodiment) FIG. 33 is a diagram showing the configuration of an image recording apparatus according to this embodiment, and a storage unit 13 is added in the third embodiment shown in FIG. The storage unit 13 is composed of a line memory, and stores the recording control signal output from the signal conversion unit 3 for one line. The output of the storage unit 13 is input to the signal conversion unit 3 and the recording density estimation unit 9.

  In this case, the signal conversion unit 3 determines the recording control signal of the target pixel with reference to the recording control signal of the pixel already converted before the target pixel output from the storage unit 13. A specific determination method will be described later. The recording control signal thus determined becomes an output signal of the control signal conversion unit 3 and is stored in the storage unit 13.

  On the other hand, the recording density estimation unit 9 estimates the recording density at the target pixel from the set of recording control signals in the vicinity of the target pixel, and outputs the recording density estimation value. The recording density estimation value output from the recording density estimation unit 9 is subtracted from the corrected pixel density value output from the adder 2 by the subtractor 7 as in the second embodiment, thereby generating an error signal. The Then, when this error signal is input to the error diffusion unit 8, error diffusion is performed on the image signal from the image input unit 1.

  The recording density estimation unit 9 in the present embodiment estimates the recording density in the vicinity of the target pixel from a set of recording control signals for the target pixel and its neighboring pixels stored in the storage unit 13 formed of a line memory. This is because the recording width (exposure pulse width) defined by the recording control signal is not necessarily proportional to the recording density, and the recording density varies depending on the positional relationship of the exposure pulses. For example, assuming that the recording control signal (exposure pulse) of the pixels around the target pixel has a distribution as shown in FIG. 34, the light distribution on the photosensitive drum 24 is as shown in FIG. From this, the recording density can be estimated by calculating the potential distribution and the toner amount distribution. Since the relationship between the exposure amount and the toner amount to be developed has a non-linear characteristic as shown in FIG. 5, the exposure pulse starts even with the same exposure pulse width of 50% as shown in FIGS. The recording density may be different at different times.

  The present embodiment is characterized in that the control is performed so that the image dots of the recorded image become a certain size. An image point is a connected area on the photosensitive drum 24 where toner is attached. As described above, when the exposure pulse width is short, the potential on the photosensitive drum 24 increases in a medium potential region, resulting in poor density stability and increased roughness noise. Therefore, in order to prevent these, the ON / OFF boundary of the exposure control signal should be minimized. However, if this boundary is made too large, the graininess becomes strong and the image quality is rather lowered. Therefore, it is necessary to control to an appropriate value.

  In particular, in the low density region, it is necessary to reduce the ratio of the dot area to the entire area. However, if one dot is reduced, the exposure amount is lowered, and the density stability is remarkably deteriorated. For this reason, even if the granularity is conspicuous to some extent in the low density region, it is necessary to increase the space between the image points so that the size of the image points does not become too small.

  In this embodiment, in order to prevent the recording width, that is, the size of the image point from becoming too small in this way, in the present embodiment, the signal conversion unit 3 refers to the recording control signal at the surrounding pixel that has already been determined as the target pixel. A pixel recording control signal is determined. FIG. 36 shows a reference range when the signal conversion unit 3 determines the recording control signal of the target pixel. In the figure, a mark X is a target pixel. The hatched portion is a pixel whose recording control signal has already been determined, and the recording control signal of these pixels is stored in the storage unit 13. In this embodiment, the recording control signal for the target pixel X is determined from the recording control signals for the adjacent pixels above and to the left of the target pixel X and the recording density value of the target pixel X. This determination procedure will be described.

  First, the provisional recording width information of the recording control signal, that is, the exposure pulse width T is determined from the recording density value of the target pixel X. Although this may be determined simply by proportional calculation, in this embodiment, taking into account the characteristics of the recording system, the non-linear relationship between the recording density on the photosensitive drum and the exposure pulse width shown in FIG. 37 is used.

  Next, as shown in Table 1, the recording control signal for the target pixel X is determined based on the recording control signals for the adjacent pixels A and B of the target pixel X. Table 1 shows recording control signals for the pixel of interest X for 16 combinations of recording control signals (recording width information = pulse width, recording position information = pulse position) for adjacent pixels A and B. FIG. 38 shows 16 recording patterns corresponding to Table 1. In FIG. 38, as in FIG. 36, the horizontal direction represents the main scanning direction and the vertical direction represents the sub-scanning direction, and the lower right pixel is the target pixel X for which the recording control signal should be determined, The adjacent pixels A and B to which the pixels are referred are represented. The hatched portion in the adjacent pixels A and B represents the recording width information and the recording position information of the recording control signal of the pixel.

表１に示す記録制御信号の決定規則の考え方について説明する。ここでは、次の３つの考え方に基づいている。第１に、注目画素Ｘの隣接画素に画点がない場合、又は左方の隣接画素に画点がなく、上方の隣接画素の上方に画点がある場合には、注目画素Ｘは孤立点となり易いので、これを防ぐために露光パルス幅を本来の露光パルス幅よりも小さくし、なおかつ露光パルス位置を後方にする。この場合、注目画素Ｘの記録濃度は本来の計算値より小さくなるため、誤差拡散処理の効果により隣接画素には大きな画点が記録され易くなる。これにより、小さい画点が形成されにくくなる。

The concept of the recording control signal determination rule shown in Table 1 will be described. Here, it is based on the following three concepts. First, if there is no image point in the adjacent pixel of the target pixel X, or if there is no image point in the left adjacent pixel and there is an image point above the upper adjacent pixel, the target pixel X is an isolated point. In order to prevent this, the exposure pulse width is made smaller than the original exposure pulse width, and the exposure pulse position is set backward. In this case, since the recording density of the pixel of interest X is smaller than the original calculated value, a large image point is easily recorded on the adjacent pixel due to the effect of the error diffusion process. This makes it difficult to form small image points.

  Second, if there is an image point in the pixel adjacent to the left of the target pixel X, the pulse position is determined so as to be connected to the image point. That is, the exposure pulse position of the target pixel X is matched with the exposure pulse position of the left adjacent pixel. However, if this exposure pulse width is narrow, it becomes isolated, so if an exposure pulse width of a certain level or more cannot be obtained, the exposure pulse width is increased or set to zero.

  Third, when there is an image point below an adjacent pixel above the target pixel X, the image point of the target pixel X is arranged in front so as to be connected to the image point. In this case, the larger the image point becomes stable, the exposure pulse width is made larger than the original value.

  Actually, since these three conditions are not exclusive, a combination of these conditions also occurs. In this case, an appropriate one is selected from the shape of the pattern. FIG. 39 shows a specific example of processing in the signal conversion unit 3. The arrangement of pixels in FIG. 39 and the meaning of the hatched portion are the same as those in FIG. 38, and the hatched portion in the target pixel X represents the recording width and recording position of the recording control signal.

  FIG. 39A shows the case where there is no image point of two adjacent pixels A and B in the upward and left directions of the target pixel X, that is, the recording control signal of the target pixel X is both 0, and the exposure pulse of the target pixel X This is a case where the provisional value of the width is T = 0.7. In this case, according to the condition 1, the recording control signal for the pixel of interest X is a signal with the exposure pulse position below and the exposure pulse width 0.2. In this way, when the image dots are not formed in the adjacent pixels A and B above and below the pixel of interest X, the image dots are formed in the lower portion of the pixel of interest X. Due to the density preservation mechanism by the diffusion process, the image point becomes large and it becomes difficult to become an isolated point.

  FIG. 39B shows a case where there is an image point in the adjacent pixel B on the left side of the target pixel X, and there is no image point in the upper adjacent pixel A, and the provisional value of the pulse width of the target pixel X is 0.5. It is. In this case, according to the condition 5, the recording control signal of the pixel of interest X is a signal with the exposure pulse position above and the exposure pulse width 0.5. As a result, the image point of the pixel of interest X is connected to the image point of the adjacent pixel on the left side, and isolation of the image point is prevented.

  FIG. 39 (c) shows a case where the reference pixel A above the target pixel X has an image point and the left adjacent pixel B has no image point, and the provisional value of the exposure pulse width of the target pixel X is 0.3. This is the case. In this case, according to the condition 3, the recording control signal for the target pixel X is a signal with the exposure pulse position above and the exposure pulse width 0.5. As a result, the image point of the target pixel X is connected to the image point of the adjacent pixel A above, and the isolation of the image point is prevented.

  As described above, in this embodiment, by determining the recording control signal of the target pixel X from the image points of the adjacent pixels A and B of the target pixel X and their positions, the connection of the image points can be controlled. It is possible to reduce the intermediate potential region on the photosensitive drum where the density tends to become unstable.

  In this embodiment, two adjacent pixels adjacent to the target pixel X are used as the reference pixels used for determining the recording control signal of the target pixel X. However, the image point of the target pixel X can be increased by taking more reference pixels. It becomes possible to control the magnitude | size of more accurately.

  As described above, according to the present embodiment, when determining the recording control signal, the image control point of the target pixel is controlled to have an appropriate size by referring to the recording control signal of the pixels around the target pixel. can do. As a result, it is possible to record an image with low graininess while improving density stability.

  (Fifth Embodiment) In the present embodiment, a scanning optical system using a laser is used in the same manner as in the first embodiment, instead of using the solid-state light emitting element arrays of the second to fourth embodiments for exposure of the photosensitive drum. An example using this will be described.

  Although overlapping with FIG. 6, FIG. 40 shows a configuration of a recording unit in the image recording apparatus according to the present embodiment. In FIG. 40, the recording control signal S1 output from the signal converter 3 is input to the pulse width converter 51. The pulse width conversion unit 51 converts a recording control signal S1 composed of recording width information and recording position information expressed by a digital signal into an exposure control signal S2 having a pulse width corresponding to the value of the signal S1. The exposure control signal S2 is a signal for controlling the light emission start time and the light emission time of the exposure light applied to the photosensitive drum. In this embodiment, the recording control signal S1 is 5 bits, of which 4 bits are a light emission time control bit and 1 bit is a light emission start time control bit. The 4-bit light emission time control bit represents a 16-valued number. When this value is x, the light emission time is given by a · x / 15. Here, a is a time width of one pixel. Further, “0” of the one-bit light emission start time control bit means that light emission starts at the beginning of one pixel, and “1” means that light emission ends at the end of one pixel.

  FIG. 41 is an example of actual signal conversion in the pulse width conversion unit 51, where (a) is a light emission start time control bit, (b) is a light emission time control bit, and (c) is an exposure control signal S2. . In the first pixel, since the value of the light emission time control bit (b) is “15”, the exposure recording control signal (c) becomes 1 during that time. In the second pixel, since the value of the light emission time control bit (b) is “7” and the light emission start time control bit (a) is “1”, only the last 7/15 period of one pixel is the exposure recording control signal ( c) is "1".

  Thus, the exposure recording control signal S2 output from the pulse width conversion unit 51 is input to the laser diode 52, and the light emitted from the laser diode 52 is controlled to be turned on / off. The light emitted from the laser diode 52 is reflected by the polygon mirror 54 via the condenser lens 53 and then focused on the photosensitive drum 24 by the imaging optical system 55 using the lens, whereby the photosensitive drum An electrostatic latent image is formed on the surface of 24. Since the polygon mirror 54 is rotating, its focal point moves parallel to the axis of the photosensitive drum 24. At this time, the f-θ lens system is used as the imaging optical system 55 so that the light is focused at any position on the axis of the photosensitive drum 24.

  The surface of the photosensitive drum 24 is made of a photosensitive material and is uniformly charged by a charger (not shown) before exposure. When light is irradiated on the surface of the photosensitive drum 24, a charge having a reverse polarity is generated at that point, and a potential distribution corresponding to the exposure amount is formed as an electrostatic latent image by canceling the charged potential. The relationship between the exposure amount of the photosensitive drum 24 and the surface potential is, for example, as shown in FIG.

  The light emission of the laser diode 52 is controlled by the exposure control signal S2. In this case, the length of one raster of the exposure control signal S2 and the rotation of the polygon mirror 54 are synchronized, and the rotation speed of the photosensitive drum 24 is adjusted to the raster interval of the exposure control signal S2, thereby electrostatic charge corresponding to the input image signal. A latent image is formed.

  Next, toner corresponding to the electrostatic potential on the photosensitive drum 24 is attached to the photosensitive drum 24 by a developing roller (not shown), and this is transferred and fixed on the recording paper, whereby an image is formed on the recording paper. . Since the configuration of the development system is the same as in FIG. 2, a detailed description thereof is omitted.

  Here, FIG. 42B shows the exposure light amount distribution on the photosensitive drum 24 when the exposure control signal S2 shown in FIG. In FIG. 42 (b), the hatched area is the exposed portion, which is a shape in which the light spot shape is convoluted with the exposure control signal of FIG. 42 (a).

  Since the exposure amount changes gently at the boundary between the exposed portion and the non-exposed portion due to the fact that the laser diode 52 is not a point light source or the blurring of the imaging optical system, on the line AA ′ in FIG. The exposure light quantity distribution is as shown in FIG. There is an intermediate exposure area near the boundary, and the area of the intermediate exposure area increases particularly in a portion where the ON / OFF change of the exposure light is large.

  Here, in the second embodiment to the fourth embodiment, the modulation direction of the recording width is the main scanning direction (direction parallel to the rotation axis of the photosensitive drum), whereas in the present embodiment, the sub scanning direction ( It differs in that it is parallel to the rotational direction of the photosensitive drum 24. Therefore, the signal conversion unit 3 outputs a recording control signal suitable for it. That is, the signal converter 3 of this embodiment is basically the same as that of the second to fourth embodiments, but the control signal determination method is different from that of the second to fourth embodiments. .

  Table 2 shows a recording control signal determination method in the signal conversion unit 3, and the positions of adjacent pixels A and B in the recording control signal determination method in the fourth embodiment shown in Table 1 are exchanged. ing.

このように、本実施形態では露光系にレーザによる走査光学系を用いた場合において、第４実施形態と同様に画点が適正な大きさになるような制御を行うことができるので、濃度安定性を高めたまま粒状性の低い画像を記録することができるという第４実施形態と同様の効果を得ることができる。

As described above, in this embodiment, when a scanning optical system using a laser is used as the exposure system, control can be performed so that the image point has an appropriate size as in the fourth embodiment. It is possible to obtain the same effect as that of the fourth embodiment in which an image with low graininess can be recorded while improving the property.

  Needless to say, the third to fifth embodiments can be applied to a color image recording apparatus as in the second embodiment. In addition, it is possible to apply the recording density estimation unit 9 and the like of the second to fourth embodiments to the first embodiment. The present invention is not limited to the embodiments of the invention described above, and it is needless to say that various modifications can be made without departing from the scope of the invention.

1 is a block diagram showing a configuration of an image recording apparatus according to an embodiment of the present invention. The block diagram which shows schematic structure of a recording part. The figure which shows the relationship between the exposure amount of a photoreceptor drum, and surface potential. FIG. 4 is a diagram illustrating a relationship between a surface potential of a photosensitive drum and a toner adhesion amount. The figure which shows the relationship between the exposure amount of a photoreceptor drum, and a toner adhesion amount. The block diagram which shows schematic structure of an exposure part. An example of a recording control pattern. The block diagram which shows an example of a structure of a 1st signal converter. The figure which shows an example of the table stored in ROM of the 1st signal conversion part. The block diagram which shows an example of a structure of a 2nd signal converter and a control signal buffer. The figure which shows an example of the table stored in ROM of a 2nd signal conversion part. The figure which shows the relationship with the adjacent pixel of a recording control pattern. The figure which shows the other example of the table stored in ROM of the 2nd signal conversion part. The figure which shows the other example of the table stored in ROM of the 2nd signal conversion part. The figure for demonstrating the operation | movement of error diffusion. The block diagram which shows the other example of a structure of a 2nd signal conversion part and an adjacent pixel control buffer. The figure which shows the other example of the table stored in ROM of the 2nd signal conversion part. The block diagram which shows the structure of the image recording apparatus which concerns on 2nd Embodiment of this invention. The block diagram which shows schematic structure of the recording part in FIG. FIG. 20 is a block diagram illustrating a configuration of a drive circuit unit in FIG. 19. The figure which shows the pixel arrangement | sequence of the image on the recording paper in 2nd Embodiment. The block diagram which shows the structure of the signal converter in FIG. 18, and a recording density estimation part. The figure which shows an example of the table stored in ROM of the signal converter in FIG. FIG. 23 is a block diagram showing a configuration of a recording density determination unit in FIG. The figure which shows the other example of the table stored in ROM of the signal conversion part in FIG. The figure which shows the fine pattern of the vertical line basic tone on a recorded image at the time of using the table shown in FIG. The figure for demonstrating the various combinations of a fine pattern. The figure which shows the image point data of the diagonal line converted into multi-value. The block diagram which shows the structure of the image recording apparatus which concerns on 3rd Embodiment of this invention. The block diagram which shows the structure of the image discrimination | determination part in FIG. The figure which shows the process algorithm for changing the recording position from the conditions of an edge part in ROM in FIG. The figure for demonstrating the change operation | movement of the recording position in 3rd Embodiment. The block diagram which shows the structure of the image recording device which concerns on 4th Embodiment of this invention. FIG. 14 is a diagram illustrating a distribution of recording control signals of pixels around a target pixel in the fourth embodiment. FIG. 35 is a view showing the distribution of exposure light on the photosensitive drum corresponding to the distribution of recording control signals in FIG. The figure for demonstrating the recording control signal determination method of the pixel of interest in the signal conversion part in FIG. The figure which shows the relationship between the recording density on a photoconductive drum, and an exposure pulse width. The figure which shows the various recording patterns in 4th Embodiment. The figure which shows the example of the specific recording pattern in embodiment. The figure which shows the structure of the image recording device which concerns on 5th Embodiment of this invention. The time chart for demonstrating operation | movement of the signal converter in FIG. 40 and a pulse width converter. The figure which shows the relationship between the exposure control signal in FIG. 40, and exposure light quantity distribution on a photosensitive drum. The figure for demonstrating the relationship between recording amount and development amount.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Image input part 2 Adder 3 (1st) signal conversion part 4 2nd signal conversion part (ROM)
DESCRIPTION OF SYMBOLS 5 Recording part 6 Control signal buffer 6a Line memory 6b Latch 7 Subtractor 8 Error diffusion part 8a Error buffer 8b Weighting part 9 Recording density estimation part 11 Image discrimination part 12 Recording position change part 13 Storage part 21 Drive circuit part 22 Exposure part ( Solid state light emitting device array)
23 Fiber lens array 24 Photosensitive drum 31 Shift register 32 Latches 33 and 34 Comparators S3 and S4 Sawtooth wave signal 37 Selector 41 Charger 42 Developer 42a Developing roller 43 Transfer roller 44 Fixing unit 45 Recording paper 46 Static discharge lamp 47 Cleaner 51 Pulse Width Conversion Unit 52 Laser Diode 53 Condensing Lens 54 Polygon Mirror 55 Imaging Optical System S1 Recording Control Signal S2 Exposure Control Signal 56 Line Memory 57 Latch D1, D2 Two Pixel Data D3 One Line Previous Data D4 Current pixel data 58 Recording density determination unit 61-64 ROM
65 Adder 66 ROM
71-74 Line memories 75-79 Multi-value / Binary determination circuit 80 Sharpness enhancement determination ROM

Claims (4)

Discrimination means for discriminating whether the input multi-value color image signal is a part of a line image or a pattern capable of improving sharpness;
The multivalued pixel density value of the target pixel of the multi-level color image signals, a signal conversion means for converting a plurality of multi-value recording control signal for defining the color of the recording width and the recording position of the pixel of interest,
When the discrimination means discriminates that the multi-value color image signal is a part of a line image or a pattern capable of improving sharpness, the recording position of each color defined by a plurality of the multi-value recording control signals is determined. When the multi-value color image signal is determined to be a part of a line image or a pattern that can improve sharpness, the recording positions of all colors defined by a plurality of multi-value recording control signals are changed independently. Change means not to change,
Recording means for multi-value recording an image based on the multi-value recording control signal changed by the changing means ;
Recording density estimation means for estimating a recording density in the recording means from the multi-value recording control signal changed by the changing means and outputting a multi-value recording density estimation value;
An image recording apparatus comprising: error diffusion means for diffusing an error between the multi-value pixel density value and the multi-value recording density estimation value into the multi-value color image signal. It said discrimination means, an image recording apparatus according to claim 1, wherein the determining the type of the multivalued color image signal to said multi-level color image signals from the multi-level quantized signal. The signal conversion means refers to a multi-value recording control signal obtained by converting multi-value pixel density values of peripheral pixels of the pixel of interest, and converts the multi-value pixel density value of the pixel into the multi-value recording control signal. The image recording apparatus according to claim 1, wherein: 2. The recording density estimation unit according to claim 1, wherein the recording density estimation unit estimates a recording density of the pixel with reference to a multi-value recording control signal obtained by converting a multi-value pixel density value of a peripheral pixel of the target pixel. The image recording apparatus according to claim 2.

Images