JP2004297267A - Pixel defect processor - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To lessen the detection error of pixel defects, even using a solid state imaging device having color filters. <P>SOLUTION: A vertical line delay unit 1 and a horizontal line delay unit 2 make coincide in time of pixels in a specified range of a picture based on image signals inputted from a solid state image sensing device 92A having color filters. A brightness signal former 3 forms brightness signals showing brightness components about the time-coincided target pixels under detection of pixel defects and adjacent pixels in the specified range. Upper, middle and lower pixel defect detectors 4-6 detect the existence of pixel defects of the target pixels, based on the brightness level difference between the target pixels and adjacent pixels, and thresholds for the white dot and black dot detections. Upper, middle and lower pixel defect correctors 7-9 correct the pixel defects, if detected. The formed brightness signals are used for detecting the pixel defects and hence the pixel defect can be exactly detected, even if the signal level of pixels for every color filter varies. <P>COPYRIGHT: (C)2005,JPO&NCIPI

Description

[0001]
TECHNICAL FIELD OF THE INVENTION
The present invention relates to a pixel defect processing apparatus that can be applied to a video camera, a vehicle-mounted camera, a surveillance camera, and the like using a color solid-state imaging device in which color filters are arranged. The present invention relates to a pixel defect processing apparatus for processing pixel defects.
[0002]
[Prior art]
2. Description of the Related Art In recent years, video cameras using a solid-state image sensor such as a CCD (Charge Coupled Device) image sensor, a CMOS (Complementary Metal Oxide Semiconductor) image sensor, a vehicle-mounted camera, and a monitoring camera have been commercialized as image pickup devices. ing.
[0003]
In the solid-state image sensors used in these devices, pixel defects occur due to manufacturing problems of the solid-state image sensors themselves, the process of manufacturing devices equipped with the solid-state image sensors, and aging after manufacturing. In some cases, the image quality may be degraded.
[0004]
Conventionally, various solutions to this problem have been proposed. When roughly classified, the coordinates at which a pixel defect has occurred are stored in a nonvolatile memory or the like in advance, and when a video is read, a defective pixel (defective pixel) is read. A defective pixel coordinate designating method (see Patent Document 1) in which the pixel in the state is replaced with a peripheral pixel, and the signal level of the target pixel is compared with the signal level of the peripheral pixel without specifying the coordinates of the defective pixel. In this case, there are two automatic pixel defect detection methods (see Patent Document 2) in which a pixel is determined to be defective and replaced with a pixel around the defective pixel.
[0005]
The defective pixel coordinate designation method can surely correct a pixel defect that has occurred up to the time of device manufacture, and is unlikely to cause side effects due to a correction error.
[0006]
However, when a large number of pixel defects occur, a non-volatile memory for storing coordinates corresponding to the number of defective pixels is required to cope with the defect. Need to be taken.
[0007]
The automatic pixel defect detection method does not require any special setting at the time of device manufacture, and can cope with a pixel defect occurring after device manufacture. Further, even if a large number of pixel defects occur, it can be dealt with.
[0008]
However, in this method, since the pixel level is detected by comparing the signal levels of the target pixel and its surrounding pixels, a determination error occurs depending on the state of the subject, and the pixel is determined to be a pixel defect even though it is not a pixel defect, On the other hand, although it is a pixel defect, it is not determined as a pixel defect, which causes deterioration in image quality.
[0009]
[Patent Document 1]
Japanese Patent Application Laid-Open No. 2001-177770 and drawings
[0010]
[Patent Document 2]
Japanese Patent Application Laid-Open No. 2002-223391 and drawings
[0011]
[Problems to be solved by the invention]
In general, when an automatic pixel defect detection method is used in a color solid-state imaging device, it is more effective to correct the pixel defect as early as possible in order to reduce the influence on others in the signal processing process.
[0012]
However, before the signal processing, the pixel data has not yet been separated for each color filter arranged in the solid-state imaging device, and the pixel data for each filter has a variation. A pixel defect cannot be determined by comparing signal levels.
[0013]
There is also a method of comparing the target pixel with pixel data of an adjacent same-color filter. However, the target pixel is separated from the target pixel by at least one pixel, and the possibility of occurrence of a determination error increases.
[0014]
In addition, if the state of the subject or the signal level of the pixel defect is low, it may not be determined that the pixel is defective.
[0015]
Further, in order to reduce the number of determination errors, it is necessary to determine a pixel defect based on the widest possible range of pixel data, but in this case, it is necessary to synchronize a plurality of horizontal lines, which requires a large number of line memories. This leads to an increase in equipment costs.
[0016]
Generally, when a signal from a solid-state image sensor is processed and converted into a video signal, a line memory for at least two horizontal lines is required for vertical edge processing and the like. Therefore, a method has been proposed in which a line memory required for video signal processing and a line memory for performing pixel defect determination are shared to reduce costs.
[0017]
When two line memories are used, pixel data for a total of three lines of current pixel data, pixel data for one horizontal line, pixel data for two horizontal lines, and pixel data for one horizontal line are aligned. Is determined as a target pixel.
[0018]
When a defective pixel enters the pixel data one horizontal line before, it is determined as a pixel defect and correction is performed. However, the problem is that the defective pixel is replaced with the current pixel data or the pixel data two horizontal lines before. If it does, it is not determined as a pixel defect.For example, when creating a vertical edge, the data of the defective pixel is used, and a false vertical edge that does not actually exist is included in the video signal, and the quality of the image is degraded I do.
[0019]
The present invention has been made to solve such a problem, and an object of the present invention is to provide a pixel defect processing apparatus that reduces a pixel error detection error even when a solid-state imaging device in which a color filter is arranged is used. That is.
[0020]
[Means for Solving the Problems]
According to an aspect of the present invention, there is provided an apparatus for processing a pixel defect by a solid-state imaging device in which a color filter is arranged, sequentially receiving a video signal from the solid-state imaging device, and synchronizing pixels in a predetermined range of an image according to the video signal. And a luminance signal generating unit that generates a luminance signal indicating a luminance component for a pixel adjacent to a target pixel for pixel defect detection in a predetermined range synchronized by the synchronizing unit, and a luminance signal generating unit that generates the luminance signal. A pixel defect detection unit that detects a pixel defect of the target pixel based on a level difference between luminance signals of the target pixel and adjacent pixels.
[0021]
According to the pixel defect processing apparatus, since the luminance signal indicating the luminance component generated by the luminance signal generating means is used for detecting the pixel defect, the solid-state imaging device in which the color filter is arranged is used to detect each pixel. Although there is a variation in the signal level of the pixel, the pixel defect can be accurately detected based on the level difference of the luminance signal.
[0022]
The above-mentioned luminance signal generating means preferably generates a luminance signal including a filter signal of each color.
[0023]
Preferably, the above-described pixel defect processing apparatus further includes a correction unit that corrects the detected pixel defect in the video signal when the pixel defect is detected by the pixel defect detection unit. Therefore, the detected pixel defect can be corrected.
[0024]
The above-mentioned synchronizing means preferably has means for sequentially inputting a video signal and synchronizing a plurality of horizontal lines, and the pixel defect detecting means is configured to detect a target pixel and each of pixels vertically and horizontally adjacent thereto. A pixel defect is detected based on the level difference between the luminance signals, and the pixels adjacent to the target pixel in each of the horizontal lines above and below the target pixel are targeted, and the luminance signals of the target pixel and the adjacent pixels are compared. A pixel defect is detected based on the level difference.
[0025]
Therefore, pixel defects can be detected and corrected for the target pixel for pixel defect detection and the pixels on the upper and lower horizontal lines.
[0026]
The video signal processed by the above-described pixel defect processing device is preferably output to a video signal processing unit prepared in advance. The video signal processing unit receives a video signal, performs at least processing based on the video signals of a plurality of horizontal lines, and outputs the processed video signal.
[0027]
Therefore, the processing based on the video signals of the plurality of horizontal lines in the video signal processing unit is performed after the pixel defect of the plurality of horizontal lines is corrected, so that the generation of noise (for example, a false vertical edge) due to the defective pixel is prevented. Can be suppressed.
[0028]
The above-mentioned synchronizing means preferably has a plurality of line memories, and the synchronizing of the plurality of horizontal lines is performed using the plurality of line memories. The processing based on the video signals of the plurality of horizontal lines in the video signal processing unit is preferably performed using a multi-line memory. Therefore, the multi-line memory can be shared for pixel defect detection and video signal processing.
[0029]
The above-mentioned luminance signal generating means preferably generates the luminance signal by adding the video signal level of the target pixel to the average value of the video signal levels of the adjacent pixels.
[0030]
Therefore, even if the pixel defect detection is performed with the peripheral pixel of the pixel defect as the target pixel, the luminance signal can be generated so as to reduce the influence of the pixel defect of the peripheral pixel. Therefore, erroneous detection of a pixel defect can be suppressed.
[0031]
The above-mentioned luminance signal generating means preferably generates a luminance signal by averaging the video signal levels of adjacent pixels.
[0032]
The above-described luminance signal generating means preferably amplifies the generated luminance signal. Therefore, even if the level of the created luminance signal is low, it can be increased to a level at which a pixel defect can be detected by amplification, and the pixel defect can be easily detected.
[0033]
The above-described luminance signal generating means preferably adds the high frequency component of the video signal of each of the target pixel and the adjacent pixel to the luminance signal generated for each of the target pixel and the adjacent pixel.
[0034]
Therefore, the high-frequency component of the created luminance signal is emphasized, so that the component of the pixel defect is emphasized, and the detection of the pixel defect becomes easy.
[0035]
The above-described luminance signal generation unit adds an offset signal to the luminance signal generated for each of the target pixel and the adjacent pixel.
[0036]
Therefore, since a noise component corresponding to the offset signal can be removed from the luminance signal, pixel defect detection based on the luminance signal can be accurately performed.
[0037]
The above-described pixel defect detection means preferably compares the level difference between the luminance signals of the target pixel and the adjacent pixels with a predetermined threshold value, and detects a pixel defect of the target pixel according to the comparison result.
[0038]
The above-mentioned threshold value is preferably changed according to the level of the video signal of the pixel around the target pixel. Since the threshold value for detecting a pixel defect is changed according to the video signal level of the peripheral pixel, the pixel defect can be detected in a state where the pixel defect is conspicuous.
[0039]
The above-mentioned threshold value is preferably increased when the level of the luminance signal of the adjacent pixel is high, and decreased when the level of the luminance signal of the adjacent pixel is low. It is decreased when the level of the luminance signal of the pixel is high, and is increased when the level is low.
[0040]
Therefore, when a defective pixel is detected, the threshold value for judging the defective pixel is changed according to the signal level of the peripheral pixel, so that the defective pixel is most conspicuous. In the case of (1), detection omission of a defective pixel can be prevented when the signal level of the peripheral pixel is high, and side effects due to erroneous determination of the defective pixel are reduced.
[0041]
The above-mentioned synchronizing means preferably has means for sequentially inputting video signals and performing synchronizing for five horizontal lines, and the pixel defect detecting means is configured to control the luminance of the target pixel and the brightness of each of the pixels adjacent vertically and horizontally. A pixel defect is detected based on a level difference between signals, and luminance signals of a pixel adjacent to the target pixel and a pixel adjacent to the target pixel on each of the horizontal lines above and below the target pixel are compared. Pixel defect is detected based on the level difference of.
[0042]
Therefore, by performing pixel defect detection for five horizontal lines simultaneously, noise in the vertical direction of the image is detected in advance, and pixel defects are corrected, so that, for example, a false vertical edge or the like occurs in video signal processing. Can be avoided. Accordingly, the same line memory can be used to perform synchronization and video signal processing for five horizontal lines for pixel defect detection.
[0043]
The above-mentioned synchronizing means preferably has means for sequentially inputting video signals and performing synchronizing for three horizontal lines, and the pixel defect detecting means is configured to control the luminance of the target pixel and the luminance of each of the vertically and horizontally adjacent pixels. A pixel defect is detected based on a level difference between signals, and a luminance signal of a target pixel and a luminance signal of a pixel adjacent to the target pixel and a pixel adjacent to the lower left and right are set for pixels adjacent to the target pixel on the horizontal line above the target pixel. A pixel defect is detected based on the level difference, and for a pixel adjacent to the target pixel on the horizontal line below the target pixel, the level difference between the luminance signal of the target pixel and each of the pixels adjacent to the upper and left and right sides is calculated. A pixel defect is detected based on this.
[0044]
Therefore, by simultaneously detecting pixel defects in three horizontal lines, noise in the vertical direction of the image is detected in advance, and by correcting the pixel defects, false vertical edges or the like are generated in the video signal processing. Can be avoided. Therefore, the same line memory can be used to perform synchronization for three horizontal lines for pixel defect detection and video signal processing.
[0045]
BEST MODE FOR CARRYING OUT THE INVENTION
Hereinafter, embodiments of the present invention will be described with reference to the drawings. In each embodiment, a pixel defect detection and correction unit mounted on an image capturing apparatus that captures an image of a subject using a field-readout color solid-state imaging device in which complementary color filters are arranged will be described.
[0046]
(Embodiment 1)
FIG. 1 shows a configuration of a pixel defect detection / correction unit according to the first embodiment, and FIG. 2 shows a configuration of an imaging device equipped with the pixel defect detection / correction unit according to each embodiment. Referring to FIG. 2, the imaging apparatus receives a video signal output from a color solid-state imaging device 92A in which a lens 91A for forming an image of a subject and a color filter are provided, and an input video signal. , A pixel defect detection / correction unit 93A (93B) that detects a pixel defect, corrects the detected pixel defect, and outputs the corrected pixel defect. A signal output from the pixel defect detection / correction unit 93A (93B) is input and processed to process a video signal. A digital signal processing circuit 94A that separates and derives a luminance signal Y and a chrominance signal C, and an encoder 95A that inputs and encodes the digital luminance signal Y and the chrominance signal C and outputs a video signal. The pixel defect detection and correction unit 93A is applied to the first embodiment, and the pixel defect detection and correction unit 93B is applied to a second embodiment described later.
[0047]
The pixel defect detection and correction unit 93A in FIG. 1 receives a signal from the color solid-state imaging device 92A and delays the signal in units of horizontal lines, a vertical line delay unit 1, a horizontal pixel delay unit 2, a luminance signal creation unit 3, and an upper pixel defect detection. It has a unit 4, a middle pixel defect detector 5, a lower pixel defect detector 6, an upper pixel defect corrector 7, a middle pixel defect corrector 8, and a lower pixel defect corrector 9.
[0048]
The vertical line delay unit 1 can use five line memories to synchronize and output signals for five horizontal lines. The horizontal pixel delay unit 2 delays the synchronized signals for five horizontal lines from the vertical line delay unit 1 in units of pixels in the horizontal direction, and outputs the delayed signals. In the present embodiment, a signal of a total of five pixels can be synchronized by delaying four pixels in the horizontal direction. The luminance signal generation unit 3 receives the synchronized signal from the horizontal pixel delay unit 2 and generates a signal indicating a luminance component of a video (hereinafter, referred to as a luminance signal) based on the input signal.
[0049]
The upper pixel defect detection unit 4 detects a pixel defect on the luminance signal created by the luminance signal creation unit 3 by comparing a pixel defect in an upper line of a pixel defect detection target pixel with peripheral pixels, and indicates a detection result. The information EH1 is output. The medium pixel defect detection unit 5 detects the pixel signal of the pixel defect detection target pixel by comparing the pixel signal of the pixel signal for the pixel defect detection with the peripheral pixels, and generates pixel defect information EH2 indicating the detection result. Output. The lower pixel defect detector 6 detects a pixel defect on the lower line of the pixel for which the pixel defect is to be detected by comparing the luminance signal created by the luminance signal generator 3 with peripheral pixels, and indicates a pixel defect indicating the detection result. The information EH3 is output.
[0050]
Upon receiving the pixel defect information EH1 from the upper pixel defect detector 4, the upper pixel defect corrector 7 estimates the signal level of the defective pixel by using the signals of the peripheral pixels of the target pixel and performs replacement. When the pixel defect information EH2 is input from the medium pixel defect detection unit 5, the medium pixel defect correction unit 8 estimates the signal level of the defective pixel using the signals of the peripheral pixels of the target pixel and performs replacement. Upon receiving the pixel defect information EH3 from the lower pixel defect detector 6, the lower pixel defect corrector 9 estimates the signal level of the defective pixel using the signals of the peripheral pixels of the target pixel and performs replacement.
[0051]
Next, the operation of the pixel defect detection and correction unit 93A according to the present embodiment will be described.
When a subject is photographed, a video signal from the color solid-state imaging device 92A is input to the vertical line delay unit 1. Here, it is assumed that the video signal from the color solid-state imaging device 92A is digitized. Although not explicitly shown in FIG. 1, in some cases, a pre-process such as a clamp process may be performed before the input to the vertical line delay unit 1 in order to facilitate the process.
[0052]
Referring to FIG. 3, vertical line delay section 1 has five line memories 101 to 105, and is a line memory capable of storing and reading video signals from color solid-state imaging device 92A for five horizontal lines. Unit 10, a memory control circuit 11 for controlling writing and reading with respect to the line memory unit 10, and a video signal read from five line memories and a video signal output from the solid-state imaging device 92A. An output data selector 12 sequentially outputs the signals as L0 to L4. FIG. 4 shows the relationship between the timing at which the video signal data input from the color solid-state imaging device 92A is written to the line memory unit 10 and the video signals output as the signals L0 to L4.
[0053]
First, writing to the line memory unit 10 will be described with reference to FIG. The video signal data H1 for one horizontal line input from the color solid-state imaging device 92A to the vertical line delay unit 1 is written to the line memory 101 as data for one horizontal line (hatched portion in FIG. 4). Video signal data H2, which is the next horizontal line, is written to the line memory 102. Similarly, video signal data H3, H4, and H5 are written into the line memories 103, 104, and 105 one line at a time in the input order. When the video signal data H6 is input, it is written into the line memory 101 again and the same operation is repeated. Reading from the line memory is performed for the four line memories except the line memory to which writing is being performed.
[0054]
By adding the video signal data input from the color solid-state imaging device 92A to the video signal data for four horizontal lines read from the line memory, the video signals for a total of five horizontal lines are synchronized. The signals are output to the horizontal pixel delay unit 2 as signals L0 to L4 corresponding to the five horizontal lines, respectively.
[0055]
In this embodiment, five line memories are used in order to synchronize video signals for five horizontal lines. However, a dual-port type line memory capable of performing writing and reading simultaneously, By performing reading and writing at double speed of the time, it is possible to synchronize five horizontal lines with four line memories.
[0056]
The horizontal pixel delay unit 2 has a circuit for synchronizing five horizontal pixels for each of the five horizontal lines of the signals L0 to L4. Here, in order to simplify the description, a configuration corresponding to one horizontal line of the horizontal pixel delay unit 2, for example, a horizontal line of the signal L0 is shown in FIG. The same configuration as that of FIG. 5 is provided for other horizontal lines.
[0057]
As shown in FIG. 5, the horizontal pixel delay section 2 delays the horizontal line of the signal L0 in the horizontal direction in units of one pixel, and has a delay element section 20 capable of synchronizing a total of five horizontal pixels. The delay element section 20 has delay elements 21 to 24. The video signal L00 first input as the signal L0 is delayed by one pixel by the delay element 21. The video signal L01 delayed by one pixel output from the delay element 21 is further input to the delay element 22, and a video signal L02 delayed by two pixels from the video signal L00 is created. The same operation is performed on the delay elements 23 and 24, and video signals L03 and L04 delayed by three and four pixels are derived, and synchronization is performed for a total of five horizontal pixels.
[0058]
The horizontal pixel delay unit 2 of the present embodiment includes the configuration shown in FIG. 5 for five horizontal lines, and a predetermined range (5 horizontal pixels in the horizontal direction X × vertical direction) of the image shown in FIG. (5 horizontal lines of Y = 25 pixels) can be synchronized. The video signal synchronized by the horizontal pixel delay unit 2 is supplied to the luminance signal creation unit 3. Here, the signals L01, L02, and L03 synchronized for the horizontal line of the signal L0 are output to the luminance signal creation unit 3. Similarly, the signals L11, L12 and L13 synchronized for the horizontal line of the signal L1, the signals L21, L22 and L23 synchronized for the horizontal line of the signal L2, the signal L31 synchronized for the horizontal line of the signal L3, The signals L41, L42, and L43 synchronized for the horizontal lines of the signals L32 and L33 and the signal L4 are also output to the luminance signal creation unit 3.
[0059]
The luminance signal generating unit 3 in FIG. 7 includes a luminance calculating unit 30 that generates a luminance signal from a synchronized video signal, an offset calculating unit that adds a negative offset to the luminance signal generated by the luminance calculating unit 30, and multiplies the gain by a gain. The apparatus includes a gain calculating section 31 and an isolated point emphasis calculating section 32 for emphasizing a horizontal isolated point based on the luminance signal subjected to the offset / gain calculation. The luminance operation unit 30 has luminance operation circuits 300 to 310, and the offset / gain operation unit 31 has offset / gain operation circuits 400 to 410 corresponding to the luminance operation circuits 300 to 310, respectively. In operation, the synchronized video signal for each horizontal line is input to each of the brightness calculation circuits of the brightness calculation unit 30, and the brightness signal is calculated.
[0060]
FIG. 8 shows a configuration of a color filter of a general complementary color type color image sensor. In FIG. 8, Cy is cyan, Ye is yellow, G is green, and Mg is magenta. FIG. 9 shows a state of a video signal output when performing field readout on an image sensor having a color filter having the configuration of FIG.
[0061]
In the complementary color type color image sensor, the filters of the four colors shown in FIG. 8 are arranged in a mosaic pattern, and when they are read out in the field, the signals of the upper and lower two filters are mixed as shown in FIG. Read as a video signal.
[0062]
In this state, it is possible to detect a pixel defect by examining the signal levels of the pixel for which a pixel defect is to be detected and its surrounding pixels, but the possibility of occurrence of a pixel defect detection error increases. This is because the light transmittances of the four color filters are different from each other, so that the absolute values of the signal levels of the video signals obtained by mixing the signals of the different filters in FIG. 9 are different.
[0063]
It is also possible to compare a pixel defect detection target pixel with a pixel in the vicinity thereof mixed with the same color filter. However, the pixel is separated from the pixel defect detection target pixel by at least one pixel width or more. Susceptible to the condition of For this reason, there is a possibility that a pixel defect is not determined as a pixel defect despite simple pixel comparison, or a pixel defect is determined even though the pixel defect is not a pixel defect. Therefore, in the present embodiment, a luminance signal indicating a luminance component is created and used.
[0064]
Generally, there are known a plurality of methods for generating a luminance signal indicating a luminance component from a signal as shown in FIG. 9, but a luminance signal is often generated from an average of pixel signals adjacent to the left and right. Signal Y indicating the luminance component to be created _yx Is the video signal level L of the pixel _yx Is given by (Equation 1).
[0065]
Y _yx = (L _yx + L _{y (x + 1)} ) ÷ 2 ... (Equation 1)
Thus, the signal Y indicating the luminance component is obtained. _yx Always include the signals of the four color filters of Cy, G, Ye, and Mg, and the absolute values of the signals become equal, so that accurate pixel defect detection becomes possible. However, in (Equation 1), if a pixel defect is present, the influence on the adjacent luminance signal is widened, so that it may be difficult to detect the pixel defect.
[0066]
Therefore, in the present embodiment, a signal indicating a luminance component (hereinafter, referred to as a luminance signal) is created by the following (Equation 2) to (Equation 4).
[0067]
Y _yx = (L _{y (x-1)} + L _{y (x + 1)} ) ÷ 2 + L _yx ... (Equation 2)
Y _yx L = (L _{(Y-1) (x-1)} + L _{(Y + 1) (x-1)} ) ÷ 2 + L _{y (x-1)} … (Equation 3)
Y _yx R = (L _{(Y-1) (x + 1)} + L _{(Y + 1) (x + 1)} ) ÷ 2 + L _{y (x + 1)} … (Equation 4)
Regarding these equations, the luminance signal Y is used for the signal determination in the vertical direction of the target pixel. _yx And the luminance signal Y for the left signal determination. _yx L signal, and the luminance signal Y _yx R signals are used respectively. In any of the equations for calculating the luminance signal, the filter signals of four colors are equally included, and the absolute values are equal, so that accurate pixel defect detection is possible. Also, compared to (Equation 1), (Equation 2) shows that the signal level L of the pixel targeted for pixel defect detection is L _yx Are added to the image to emphasize it, which makes it easier to detect a defective pixel. In the present embodiment, the processing using the luminance signal of (Equation 2) will be described, but the same processing can be performed even when the luminance signal of (Equation 1) is applied instead.
[0068]
Since the video signal corresponding to the pixel defect contains many high frequency components due to the pixel defect, a high frequency component extracted from the video signal corresponding to the created luminance signal of the pixel may be added. Thereby, the signal component due to the pixel defect is emphasized in the created luminance signal, so that the detection of the defective pixel based on the luminance signal becomes easy.
[0069]
In the present embodiment, the luminance calculation circuits 300 to 304 in the luminance calculation unit 30 generate the luminance signals Y0, Y1, Y2, Y3, and Y4 using (Equation 2), and generate the luminance calculation circuits 305, 307, and 309. Creates the luminance signals Y1L, Y2L and Y3L using (Equation 3), and the luminance calculation circuits 306, 308 and 310 create the luminance signals Y1R, Y2R and Y3R using (Equation 4). The luminance signals generated by the luminance operation circuits 300 to 310 are input to the corresponding offset / gain operation circuits 400 to 410, respectively.
[0070]
Each of the offset / gain calculation circuits 400 to 410 adds a preset negative offset and multiplies the input luminance signal by a preset gain in order to emphasize a pixel defect portion. I do.
[0071]
By performing such an operation, when detecting a white point when a subject in which a pixel defect is most conspicuous is dark, a signal having a small signal level is truncated by adding a negative offset and applying a gain to detect a defective pixel. This has the effect of amplifying the signal level of the pixel signal to more reliably detect pixel defects. As the offset value and the gain value, values that are considered to be optimal depending on the noise generation state and usage state are set for each system.
[0072]
The luminance signal output from the offset / gain calculating unit 31 is input to the isolated point emphasizing operation circuits 321 to 323 of the isolated point emphasizing operation unit 32, and the arithmetic operation for the isolated point emphasis is performed and output. Here, according to (Equation 5) or (Equation 6), the emphasis component (YL + YR) is added to emphasize the isolated point.
[0073]
EY = Y- (YL + YR) (Equation 5)
YE = Y−G × (YL + YR) (Equation 6)
FIG. 10 shows a state in which an isolated point is emphasized according to (Equation 5). In FIG. 10, the luminance signal Y of the pixel defect detection target pixel, the luminance signal YL of the pixel one pixel left adjacent to the pixel defect detection target pixel, the luminance signal YR of the pixel one pixel right from the pixel defect detection target pixel, and the isolated signal The luminance signal YE after performing the point emphasis is shown. When the isolated point enhancement is performed, the pixel defect (pixel defect E1, pixel defect E2) becomes the emphasized pixel defect (pixel defect EE1, pixel defect EE2) as shown in FIG. 10, and the detection of the pixel defect becomes easy.
[0074]
Further, the isolated component may be emphasized by multiplying the emphasis component by the gain G according to (Equation 6). According to (Equation 6), the pixel defect is further emphasized and the detection of the pixel defect becomes easier. However, when the video signal contains much noise, the noise component is emphasized and is erroneously determined as a defective pixel. Maybe. Therefore, it is preferable to set the value of the gain G according to the state of the system.
[0075]
Among the luminance signals generated by the luminance signal generating unit 3, the luminance signals Y1, Y0, Y2, Y1L and Y1R are given to the upper pixel detecting unit 4, and the luminance signals Y2, Y1, Y3, Y2L and Y2R are medium pixels. The luminance signals Y3, Y2, Y4, Y3L, and Y3R are supplied to the detection unit 5, and are supplied to the lower pixel defect detection unit 6. FIG. 11 shows the configuration of the upper pixel defect detector 4. Since each of the middle pixel defect detector 5 and the lower pixel defect detector 6 has the same configuration as the upper pixel defect detector 4, only the configuration of the upper pixel defect detector 4 will be described here. The description of the lower pixel defect detector 6 is omitted.
[0076]
In FIG. 11, an upper pixel defect detection unit 4 includes a white point threshold control unit 40 that determines a threshold value K1 for detecting a pixel defect (white point) based on the signal level of a pixel around a pixel defect detection target pixel, a white point investigation unit 41, A black point threshold control unit 43 that determines a threshold value K2 for detecting a pixel defect (black point) based on a signal level of a pixel surrounding the pixel to be detected; and a pixel defect detection pixel based on a signal from the white point inspection unit 41 Is provided with a defective pixel determination unit 44 for determining whether or not a pixel is a defective pixel, and a defective pixel determination unit 45 for determining whether or not a pixel defect detection target pixel is a defective pixel based on a signal from the black spot investigation unit 42. The defective pixel determination units 44 and 45 output pixel defect information EH1 when determining that the target pixel is a defective pixel.
[0077]
When detecting a white point, the white point investigating unit 41 compares the signal levels of the detection target pixel and the pixels adjacent in the vertical and horizontal directions using the threshold value K1. Includes white point right direction checking circuits 411, 412, 413 and 414. When detecting a black point, the black point investigating unit 42 compares a signal level of a pixel to be detected with a pixel adjacent in the vertical and horizontal directions using a threshold value K2. 421, 422, 423 and 424.
[0078]
The upper pixel defect detection unit 4 compares the level of the luminance signal of the pixel defect detection target pixel and the level of the luminance signal of the pixel adjacent in the vertical and horizontal directions, and determines that the luminance signal level of the pixel defect detection target pixel is that of the adjacent pixel. If it is more protruding and isolated, the target pixel is determined to be a defective pixel. Specifically, when white point detection is performed, it is determined that a pixel is defective if all of the conditions of (Expression 7) to (Expression 10) are satisfied.
[0079]
Y _y -Y _(Y-1) > K1 (Equation 7), Y _y -Y _{(Y + 1)} > K1 (Equation 8)
Y _y -Y _y L> K1 (Equation 9), Y _y -Y _y R> K1 (Equation 10)
(Equation 7) is an upward inspection of the pixel for pixel defect detection, (Equation 8) is an downward inspection, (Equation 9) is a leftward inspection, and (Equation 10) is a rightward inspection. Each has gone. When all of the conditions of (Equation 7) to (Equation 10) are satisfied, it means that the pixel defect detection target pixel has a higher signal level than the adjacent pixel, that is, is isolated, and the pixel defect detection target pixel is isolated. ) And pixel defect information EH1 is output.
[0080]
Similarly, when black point detection is performed, a pixel defect is determined if all of the conditions of (Equation 11) to (Equation 14) are satisfied.
[0081]
Y _(Y-1) -Y _y > K2 (Equation 11), Y _{(Y + 1)} -Y _y > K2 (Equation 12)
Y _y LY _y > K2 (Equation 13), Y _y RY _y > K2 (Equation 14)
(Equation 11) is an upward inspection of the pixel to be detected, (Equation 12) is an downward inspection, (Equation 13) is a leftward inspection, and (Equation 14) is a rightward inspection. Each has gone. The satisfaction of all of the conditions of (Equation 11) to (Equation 14) means that the pixel defect detection target pixel has a lower signal level than the adjacent pixel, that is, is isolated, and the pixel defect (black dot) ) And the pixel defect information EH1 is output.
[0082]
The thresholds K1 and K2 may be fixed values according to the state of the system, or may be variably set according to the signal levels of adjacent peripheral pixels in order to further prevent omission of pixel defect detection.
[0083]
For example, the average value of the luminance signal levels of the pixels in the vertical and horizontal directions of the pixel defect detection target pixel is calculated, and if the average value of the luminance signal levels is high, the threshold K1 is increased and the threshold K2 is decreased, and conversely, If the average value of the luminance signal levels is low, control is performed so that the threshold value K1 is reduced and the threshold value K2 is increased.
[0084]
By controlling the thresholds K1 and K2 in this way, even if the signal level of the peripheral pixel of the target pixel is high (low) and the target pixel is a white (black) point in the case of detecting a white (black) point, It is possible to easily detect a white (black) point and prevent image quality deterioration due to erroneous determination.
[0085]
Further, when the signal level of the peripheral pixel is low, the presence of a white point is noticeable, so that the threshold value K1 is reduced so as to be surely detected, thereby preventing the white point from being missed.
Similarly, in the case of black point detection, if the signal level of the peripheral pixel is low, the black point is present even if it is not detected, and the black point is not conspicuous even if it is not detected. Therefore, the threshold value K2 is increased to prevent image quality deterioration due to erroneous determination. In addition, when the signal level of the peripheral pixel is high, the presence of a black point is noticeable, so the threshold value K2 is set small so as to be surely detected, thereby preventing the black point from being missed.
[0086]
Further, in the present embodiment, the detection of pixel defects is performed simultaneously on the signals L1, L2 and L3 of three horizontal lines from the video signals of the signals L0 to L4 of the five horizontal lines. This is because performing processing using signals of a plurality of horizontal lines, such as vertical edge processing, after performing pixel defect correction on only the signal L2, causes a phenomenon that adversely affects image quality, such as false vertical edges. is there.
[0087]
Specifically, when the digital signal processing circuit 94A creates a luminance signal using the video signal of the signal L2, the vertical edge processing is generally performed using the video signals of the signals L1 and L3. This is because, if only the video signal of the signal L2 is corrected, the pixel defect of the luminance signal itself is corrected, but if the pixel defect exists in the horizontal line of the signal L1 or L3, the defective pixel is used for the vertical edge processing. This is because a false vertical edge that should not exist originally appears.
[0088]
In this embodiment, the upper, middle, and lower pixel defect detectors 4, 5, and 6 detect pixel defects for all three horizontal line signals L1, L2, and L3, and the pixel defect corrector 7 corrects them. Therefore, phenomena such as false vertical edges do not occur.
[0089]
When a pixel defect is detected by the upper pixel defect detector 4, the middle pixel defect detector 5, and the lower pixel defect detector 6, the pixel defect information EH1, EH2, and EH3 are respectively transmitted to the upper pixel defect corrector 7, the middle pixel Since the image data is output to the defect correction unit 8 and the lower pixel defect correction unit 9, the upper pixel defect correction unit 7 receives the pixel defect information EH1, the middle pixel defect correction unit 8 receives the pixel defect information EH2, and When the pixel defect information EH3 is input, the pixel defect correction unit 9 corrects each pixel defect.
[0090]
FIG. 12 shows the configuration of the upper pixel defect correction unit 7. Since the middle pixel defect correction unit 8 and the lower pixel defect correction unit 9 have the same configuration as the upper pixel defect correction unit 7, only the configuration of the upper pixel defect correction unit 7 will be described here. The description of the pixel defect correction unit 7 is omitted.
[0091]
In FIG. 12, the upper pixel defect correction unit 7 receives the signals L10 and L14 from the horizontal pixel delay unit 2, generates a correction signal for the defective pixel based on the average value, and outputs the correction signal, and a selector 51. Having. The selector 51 receives the correction signal from the correction value calculation unit 50 and the signal L12 (uncorrected signal) from the horizontal pixel delay unit 2 and inputs when the pixel defect information EH1 is given from the upper pixel defect detection unit 4. A correction signal among the signals is derived, and otherwise, a signal L12 (uncorrected signal) is derived.
[0092]
The correction value calculation unit 50 calculates an average value of the signals L10 and L14 of the same color filter as the signal L12 which is the pixel to be subjected to the pixel defect detection among the data of the five horizontal pixels synchronized by the horizontal pixel delay unit 2. By doing so, a correction signal of the signal L12 is created (see FIGS. 6 and 8). This is based on the fact that the signal L12 of the defective pixel to be detected and the signals L10 and L14 of the same color filter adjacent on the left and right in the video signal generally have a strong correlation.
[0093]
The correction signal created here is input to the selector 51 together with the signal L12 which is an uncorrected signal. The selector 51 outputs the signal L12 when the pixel defect information EH1 is not supplied from the upper pixel defect detector 4 and indicates that the pixel is not a pixel defect, and the selector 51 is supplied with the pixel defect information EH1 and indicates that the pixel is defective. In this case, the correction of the pixel defect is performed by outputting the correction signal created by the correction value calculation unit 58.
[0094]
The upper pixel defect corrector 7 performs defective pixel correction on the horizontal line of the signal L1. Similarly, the middle pixel defect corrector 8 corrects the horizontal line of the signal L2, and the lower pixel defect corrector 9 corrects the horizontal line of the signal L3. Correction of defective pixels is performed simultaneously.
[0095]
(Embodiment 2)
In the pixel defect detection and correction unit 93A of the first embodiment, the vertical line delay unit 1 uses the line memories of five horizontal lines and synchronizes the video signals of five horizontal lines. The second pixel defect detection and correction unit 93B detects and corrects pixel defects using line memories for three horizontal lines.
[0096]
Referring to FIG. 13, pixel defect detection and correction section 93B delays a digital video signal from solid-state color image sensor 92A in units of horizontal lines and outputs the delayed signal, and outputs from vertical line delay section 60. A horizontal pixel delay unit 61 that delays and outputs the signals for three horizontal lines on a pixel-by-pixel basis in the horizontal direction; a luminance signal generation unit 62 that generates and outputs a luminance signal from the signal from the horizontal pixel delay unit 61; An upper pixel defect detector 63, a middle pixel defect detector 64, a lower pixel defect detector 65, an upper pixel defect corrector 66, a middle pixel defect corrector 67, and a lower pixel defect corrector 68 are provided.
[0097]
The vertical line delay unit 60 uses line memories for three horizontal lines, and can synchronize signals for three horizontal lines. The horizontal pixel delay section 61 delays by four pixels in the horizontal direction, so that signals for a total of five horizontal pixels can be synchronized.
[0098]
The upper pixel defect detection unit 63 detects a pixel defect in the upper line of the pixel defect detection target pixel from the luminance signal generated by the luminance signal generation unit 62 by comparing the pixel defect with a peripheral pixel. The medium pixel defect detection unit 64 detects a pixel defect of the pixel defect detection target pixel by comparing the pixel defect with a peripheral pixel from the luminance signal generated by the luminance signal generation unit 62. The lower pixel defect detection unit 65 detects a pixel defect on the lower line of the pixel to be detected from the luminance signal generated by the luminance signal generation unit 62 by comparing the pixel defect with a peripheral pixel.
[0099]
The upper pixel defect correction unit 66 estimates the signal level of the defective pixel based on the pixel defect information detected by the upper pixel defect detection unit 63 using the signals of the pixels surrounding the pixel targeted for pixel defect detection, and replaces the signal level. The medium pixel defect correction unit 67 estimates the signal level of the defective pixel based on the pixel defect information detected by the medium pixel defect detection unit 64 using the signals of the peripheral pixels of the pixel targeted for the pixel defect detection, and replaces the signal level. The lower pixel defect corrector 68 estimates the signal level of the defective pixel based on the pixel defect information detected by the lower pixel defect corrector 65 by using the signals of the peripheral pixels of the pixel targeted for the pixel defect detection, and replaces it.
[0100]
Referring to FIG. 14, vertical line delay section 60 includes three line memories 701, 702, and 703, and stores video signals input from color solid-state imaging device 92A for three horizontal lines by line memories 701, 702, and 703. A line memory unit 70 capable of reading and reading, a memory control circuit unit 71 for controlling writing and reading of the three line memories, and a video signal read from the three line memories from the signals L0 to L2. And an output data selector 72 for selecting and outputting the data in the order shown in FIG.
[0101]
FIG. 15 shows the relationship between the timing at which the data of the video signal input from the color solid-state imaging device 92A is written to the line memory unit 70 and the signal L0, L1, and L2.
[0102]
Referring to FIG. 15, data H1 of a video signal for one horizontal line input from color solid-state imaging device 92A to vertical line delay unit 60 is written to line memory 701 as data for one horizontal line. The video signal data H2, which is the next horizontal line, is written into the line memory 702. Further, the video signal data H3 of the next horizontal line is written to the line memory 703.
[0103]
Further, when the video signal data H4 is inputted next, the data is written into the line memory 701 again, and thereafter, every time the video signal data is inputted, the writing into the line memory is performed. Reading from each of the line memories 701 to 703 is performed on two line memories to which writing has not been performed.
[0104]
By adding video signal data input from the color solid-state imaging device 92A to video signal data for two horizontal lines read from the line memory, video signals for a total of three horizontal lines are synchronized, and an output data selector 72 is output. Are output as signals L0, L1 and L2.
[0105]
In the second embodiment, three line memories 701 to 702 are used in order to synchronize video signals for three horizontal lines. However, a dual port type line memory capable of performing writing and reading at the same time, By performing reading and writing at twice the time per pixel, it is possible to synchronize three horizontal lines with two line memories.
[0106]
The video signal data for three horizontal lines synchronized by the vertical line delay unit 60 is given to the horizontal pixel delay unit 61. The operation of the horizontal pixel delay unit 61 is the same as the operation of the horizontal pixel delay unit 2 described in the first embodiment, and a description thereof will be omitted. The video signal data synchronized by the horizontal pixel delay unit 61 is provided to the luminance signal creation unit 62.
[0107]
Referring to FIG. 16, a luminance signal creating unit 62 creates a luminance signal from a synchronized video signal, adds a negative offset from the luminance signal created by the luminance computing unit 80, and adjusts the gain. An offset / gain operation unit 81 to be applied and an isolated point emphasis operation unit 82 for emphasizing a horizontal isolated point from the luminance signal subjected to the offset / gain operation are provided. Luminance operation section 80 has luminance operation circuits 800 to 808, and offset / gain operation section 81 has offset / gain operation circuits 900 to 908 corresponding to luminance operation circuits 800 to 808, respectively. In operation, the synchronized video signal for each horizontal line is input to each of the brightness calculation circuits of the brightness calculation unit 80, and the brightness signal is calculated.
[0108]
In the second embodiment, a signal indicating a luminance component is created by the following three types of equations.
Y _yx = (L _{y (x-1)} + L _{y (x + 1)} ) ÷ 2 + L _yx … (Equation 15)
Y _yx L = (L _{(Y-1) (x-2)} + L _{(Y + 1) (x-1)} ) ÷ 2 ... (Equation 16)
Y _yx R = (L _{(Y-1) (x + 1)} + L _{(Y + 1) (x + 2)} ) ÷ 2 ... (Equation 17)
The signal Y is used to determine the signal in the vertical direction of the target pixel. _yx And the signal Y _yx L and the signal Y _yx R is used for each. In any of the above equations, the filter signals of the four colors are equally included, and the absolute values are equal, so that accurate pixel defect detection is possible. Created signal Y _yx L, signal Y _yx R and signal Y _yx Is hereinafter referred to as a luminance signal.
[0109]
As in the first embodiment, the created luminance signal is added with a negative offset by an offset / gain calculator 81 and multiplied by a gain, and is further subjected to an isolated point emphasis calculation by an isolated point emphasis calculator 82. And is provided to a pixel defect detection unit.
[0110]
FIG. 17 shows the configuration of the upper pixel defect detector 63, and FIG. 18 shows the configuration of the lower pixel defect detector 65. The configuration of the middle pixel defect detector 64 is the same as that of the middle pixel defect detector 5 of the first embodiment (see FIG. 11).
[0111]
Referring to FIG. 17, upper pixel defect detection unit 63 determines a detection threshold K1 of a defective pixel (white point) from a peripheral signal level of a pixel to be detected, and a white point threshold control unit 90. A white point investigator 91 that compares the signal level of the pixel adjacent to the target pixel in the lower left and right direction; a black point inspector 92 that compares the signal level of the target pixel and the pixel adjacent in the lower left and right direction when detecting the black point; A black point threshold control unit 93 that determines a detection threshold value K2 of a defective pixel (black point) based on the signal level of a peripheral pixel of the detection target pixel, and whether a pixel defect detection target pixel is a defective pixel is determined by a signal from a white point examination unit 91. A defective pixel determination (white point) unit 94 for performing determination and a defective pixel determination (black point) unit 95 for determining whether a pixel targeted for pixel defect detection is a defective pixel based on a signal from the black point investigation unit 92 are provided.
[0112]
Referring to FIG. 18, lower pixel defect detection unit 65 determines a detection threshold K1 of a defective pixel (white point) from a signal level of a peripheral pixel of a defective pixel detection target pixel. A white point investigating unit 101 that compares the signal level of the pixel to be detected with the pixel vertically adjacent to the upper left and right direction, a black point inspecting unit 102 that compares the signal level of the pixel to be detected with the pixel adjacent to the upper right and left direction when detecting the black point Whether the pixel defect detection target pixel is a defective pixel based on a signal from the black point threshold control unit 103 and the white point inspection unit 101 that determines the detection threshold K2 of the defective pixel (black point) from the signal level of the peripheral pixel of the pixel defect detection target pixel A defective pixel determination (white point) unit 104 for determining whether or not a pixel defect detection target pixel is a defective pixel based on a signal from the black pixel inspection unit 102 and a defective pixel determination (black point) It comprises a section 105.
[0113]
In the first embodiment, based on the synchronized signals for five horizontal lines, the pixel defect is determined by comparing the signal levels of the pixel defect detection target pixel and its adjacent pixels in the vertical and horizontal directions. However, in the second embodiment, since only three horizontal lines are synchronized, the pixel defect detection of the horizontal line of the signal L0 is performed by the signals of the three pixels in the lower left and right directions adjacent to the pixel to be detected. For detecting a pixel defect in the horizontal line of the signal L2, the pixel defect is detected based on the signal levels of pixels in the three directions in the upper right and left directions adjacent to the pixel to be detected.
[0114]
As for the pixel defect detection of the horizontal line of the signal L1, the pixel defect is detected from the signal levels of the four pixels in the four directions (up, down, left, and right) adjacent to the pixel to be detected as in the first embodiment.
[0115]
Since the pixel defect detection of the horizontal line of the signal L0 and the signal L1 is performed only with the signal of the pixel in the three directions adjacent to the pixel defect detection target pixel, it is difficult to obtain high detection accuracy as compared with the first embodiment. . However, the cost can be reduced because the line memory is sufficient for three lines. Therefore, when configuring the system, one of the configurations of the first and second embodiments may be selected depending on whether image quality or cost is important.
[0116]
When the upper pixel defect detector 63, the middle pixel defect detector 64, and the lower pixel defect detector 65 detect a pixel defect, the pixel defect information EH0, EH1, and EH2 are respectively converted to an upper pixel defect corrector 66, a middle pixel defect corrector 67, and This is given to the lower pixel defect correction unit 68. The pixel defect correction performed by the upper pixel defect corrector 66, the middle pixel defect corrector 67, and the lower pixel defect corrector 68 in response to each of the pixel defect information EH0, EH1, and EH2 is described in the embodiment. Description is omitted because it is similar to 1.
[0117]
The embodiments disclosed this time are to be considered in all respects as illustrative and not restrictive. The scope of the present invention is defined by the terms of the claims, rather than the description above, and is intended to include any modifications within the scope and meaning equivalent to the terms of the claims.
[0118]
【The invention's effect】
According to the present invention, since a luminance signal indicating a luminance component generated by the luminance signal generating means is used for detecting a pixel defect, a pixel for each color filter is used by using a solid-state imaging device in which a color filter is arranged. Despite the variation in the signal level, the pixel defect can be accurately detected based on the level difference of the luminance signal.
[Brief description of the drawings]
FIG. 1 is a configuration diagram of a pixel defect detection and correction unit according to a first embodiment.
FIG. 2 is a configuration diagram of an imaging device applied to each embodiment.
FIG. 3 is a configuration diagram of a vertical line delay unit 1;
FIG. 4 is a diagram illustrating a relationship between timings at which signal data input from an image sensor is written to a line memory and signals L0 to L4 read out;
FIG. 5 is a configuration diagram of a part (for one horizontal line) of a horizontal pixel delay unit 2;
FIG. 6 is a diagram showing a range synchronized by a vertical delay unit 1 and a horizontal pixel delay unit 2;
FIG. 7 is a configuration diagram of a luminance signal creation unit 3.
FIG. 8 is a configuration diagram of a color filter of a general complementary color type color imaging device.
FIG. 9 is a diagram showing a video signal output when field reading is performed.
FIG. 10 is a diagram illustrating an isolated point emphasis calculation.
FIG. 11 is a configuration diagram of an upper pixel defect detection unit 4;
FIG. 12 is a configuration diagram of an upper pixel defect correction unit 7;
FIG. 13 is a configuration diagram of a pixel defect detection and correction unit according to the second embodiment.
14 is a configuration diagram of a vertical line delay unit 60. FIG.
FIG. 15 is a diagram illustrating a relationship between timings at which signal data input from an imaging element according to the second embodiment is written to a line memory and signals L0 to L2.
FIG. 16 is a configuration diagram of a luminance signal creation unit 62.
FIG. 17 is a configuration diagram of an upper pixel defect detection unit 63;
FIG. 18 is a configuration diagram of a lower pixel defect detection unit 65;
[Explanation of symbols]
1,60 vertical line delay unit, 2,61 horizontal pixel delay unit, 3,62 luminance signal creation unit, 4,63 upper pixel defect detector, 5,64 middle pixel defect detector, 6,65 lower pixel defect detector , 7, 66 upper pixel defect corrector, 8, 67 middle pixel defect corrector, 9, 68 lower pixel defect corrector, 10, 70 line memory unit, 30 luminance calculator, 31, 81 offset / gain calculator, 32 , 82 Isolated point emphasis calculation section, 93A, 93B Pixel defect detection and correction section, K1, K2 threshold value.

Claims (11)

An apparatus for processing a pixel defect by a solid-state imaging device in which a color filter is arranged,
A video signal is sequentially input from the solid-state imaging device, and a synchronization unit that synchronizes pixels in a predetermined range of an image according to the video signal,
A luminance signal generation unit that generates a luminance signal indicating a luminance component for a target pixel and an adjacent pixel in the pixel defect detection in the predetermined range synchronized by the synchronization unit;
A pixel defect detection unit configured to detect a pixel defect of the target pixel based on a level difference between the luminance signals of the target pixel and the adjacent pixels generated by the luminance signal generation unit. . 2. The pixel defect processing apparatus according to claim 1, further comprising: a correction unit configured to correct the detected pixel defect in the video signal when the pixel defect is detected by the pixel defect detection unit. The luminance signal creating means,
3. The pixel defect processing according to claim 1, wherein the luminance signal is created by adding the video signal level of the target pixel to an average value of the video signal levels of the adjacent pixels. 4. apparatus. The luminance signal creating means,
The pixel defect processing apparatus according to any one of claims 1 to 3, wherein the created luminance signal is amplified. The luminance signal creating means,
4. The method according to claim 1, wherein a high frequency component of the video signal of each of the target pixel and the adjacent pixel is added to the luminance signal created for each of the target pixel and the adjacent pixel. The pixel defect processing apparatus according to claim 1. The luminance signal creating means,
The pixel defect processing apparatus according to any one of claims 1 to 5, wherein an offset signal is added to the luminance signal created for each of the target pixel and the adjacent pixel. The pixel defect detection means,
7. The method according to claim 1, wherein a level difference between the luminance signals of the target pixel and the adjacent pixels is compared with a predetermined threshold value, and a pixel defect of the target pixel is detected according to a comparison result. A pixel defect processing apparatus according to any one of the above. The pixel defect processing apparatus according to claim 7, wherein the threshold value is changed according to a level of the video signal of a pixel around the target pixel. When detecting the white point of the target pixel, when the level of the luminance signal of the adjacent pixel is high, the threshold is increased, and when the level is low, the threshold is decreased, and when detecting the black point of the target pixel, 9. The pixel defect processing apparatus according to claim 8, wherein the threshold value is reduced when the level of the luminance signal of the adjacent pixel is high, and the threshold value is increased when the level is low. The synchronization means has means for sequentially inputting the video signal and performing synchronization for five horizontal lines,
The pixel defect detection means,
A pixel defect is detected based on a level difference between the luminance signals of the target pixel and each of pixels adjacent to the top, bottom, left, and right,
For a pixel adjacent to the target pixel in each of the horizontal lines above and below the target pixel, a pixel defect is detected based on a level difference between the luminance signals of the target pixel and each of the pixels adjacent to the upper, lower, left, and right sides. The pixel defect processing apparatus according to claim 1, wherein: The synchronization means includes means for sequentially inputting the video signal and performing synchronization for three horizontal lines,
The pixel defect detection means,
A pixel defect is detected based on a level difference between the luminance signals of the target pixel and each of pixels adjacent to the top, bottom, left, and right,
For a pixel adjacent to the target pixel of the horizontal line above the target pixel, a pixel defect is detected based on a level difference between the luminance signals of the target pixel and pixels adjacent to the lower left and right,
For a pixel adjacent to the target pixel on the horizontal line below the target pixel as a target, detecting a pixel defect based on a level difference between the luminance signals of the target pixel and pixels adjacent to the upper and lower left and right, respectively. The pixel defect processing apparatus according to claim 1, wherein:

Images