JPH0773336B2 - Image processing device - Google Patents

Description

Description: TECHNICAL FIELD The present invention relates to an image processing apparatus having an image identification function suitable for use in a digital copying machine, facsimile, or the like.

[Prior art]

In this type of device, a device for identifying the image tone (characteristic or characteristic of the image) of the pixel of interest by utilizing the density difference in the vicinity of the pixel of interest, or a binarized image in an area (block) of about 4 × 4. An apparatus or the like that divides and identifies the image tone for each block based on the dot arrangement in the block can be considered. However, the former has to perform a calculation of about 6 bits and cannot be realized at a low cost. Further, the latter has a drawback that the accuracy is poor because of the 1-bit judgment.

〔Purpose〕

The present invention has been made in view of the above points, and an object of the present invention is to provide an image processing apparatus capable of accurately identifying image features with a relatively simple configuration.

〔Example〕

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

FIG. 1 shows four neighboring pixels including the target pixel A. An image identifying method in this embodiment will be described with reference to FIG. In this embodiment, the image tone is identified for each pixel by using the upper 2 bits of the quantized image data of 6 bits. The maximum density Lmax and the minimum density Lmin are calculated in the pixels B, C, D and E adjacent to the target pixel A, and the density gradient ΔL
To calculate. (1) When ΔL = Lmax−Lmin ≧ 2 (when ΔL = 2 or 3) ... The pixel of interest belongs to the edge portion of a line image such as a character.

(2) When ΔL = Lmax−Lmin <2 (when ΔL = 0 or 1) ... The pixel of interest belongs to the halftone region of a photograph or the like.

If the pixel of interest is determined to be an edge portion such as a character, the pixel of interest is binarized with a fixed threshold value, and if it is determined to be a halftone image such as a photograph, the pixel of interest is binarized by dither processing. It will be transformed.

If the character part is sharply binarized even in one or two pixels in the one-dimensional direction in the edge part, the quality of the reproduced character is remarkably improved even if the inside of the character is dithered (halftone processing). . However, the edge portion must be binarized two-dimensionally.

In this embodiment, the edge portion of the character portion is two-dimensionally identified by using the above-described identification method, and a high-quality character is reproduced.

FIG. 2 is an example of a circuit diagram of an image processing apparatus for binarizing image data and executing the above-mentioned discrimination, and the numerical value on the line indicates the bit number of the image data. In the figure, 1 is a solid-state image sensor (sensor) 1a and sensor 1 such as CCD.
An original reading unit having an A / D converter 1b for quantizing the output from a into 6-bit image data. The document reading unit has a line sensor to electrically scan the document in the line direction (main scanning direction) and mechanically scan the document in the direction perpendicular to the line direction (sub scanning direction) to read the entire document. It is a thing. Reference numeral 3 is a simple binarization circuit that compares the 6-bit image data sequentially input with a predetermined threshold value (fixed threshold value) and outputs binarized signals "1" and "0". To explain further, the simple binarization circuit 3 sequentially outputs binarization signals along the main scanning direction X, and, for example, when the output of the first line 1 is completed, the second line l2 Start output. Reference numeral 4 is a dither processing circuit that compares the 6-bit image data that is sequentially input with each threshold value of a predetermined dither matrix and outputs halftone-processed binarized data. The operation of the digitizing circuit 3 is performed in parallel.

Reference numeral 7 is a delay RAM for delaying the binarized signal output from the simple binarization circuit 3 by one line in the sub-scanning direction. Similarly, 8 is the binarized signal output from the dither processing circuit 4 for one line. It is a delay RAM that delays. Reference numerals 15 and 16 denote delay flip-flop circuits (DF / F circuits) which delay the binarized outputs from the delay RAMs 7 and 8 by one pixel in the main scanning direction.

Reference numerals 5 and 6 are delay RAMs for inputting the upper 2 bits of the 6-bit image data output from the document reading unit 1 and delaying them by one line in the sub-scanning direction. Reference numerals 9 to 12 are D to F / F circuits. 2 bit image data from the delay RAMs 5 and 6 in the main scanning direction
Delay by pixel.

In this embodiment, the delay RAMs 5 and 6 and the D-F / F circuits 9 to 1 are used.
2. A discrimination circuit is constructed by using the ROM 13 to discriminate image tone for each pixel of interest.

Now, the pixel D adjacent to the target pixel A shown in FIG.
If it is output from the F circuit 9, R delayed by one line
The pixel E is delayed from the AM5 by the pixel E, the output of the RAM5 is delayed by one pixel, and the target pixel A is delayed from the D-F / F circuit 10, and the output of the D-F / F circuit 10 is further delayed by one pixel D-. From the F / F circuit 12, the pixel C and the output of RAM5 are further delayed by one line in RAM6, and RA
The pixel B is output from the D-F / F circuit 11 in which the output of M6 is delayed by one pixel. With the above configuration, the 2-bit density data of the pixel B, C, D and E adjacent to the pixel of interest A can be detected simultaneously. The pixel data A to E are input to the address terminals A0 to A9 of the ROM (memory) 13.

An identification result of "0" or "1" is stored in advance in the ROM 13 according to the address, and it is determined whether the target pixel A is an edge portion such as a character or a halftone image such as a photograph. It is configured to be able to do.

That is, the ROM 13 has 2 bits (0 to 5 pixels) for each of 5 pixels A to E.
It has an address corresponding to all cases that change in 3), and stores the identification result of "0" or "1" corresponding to all the addresses.

Therefore, in the ROM 13, the same processing as the above-described calculation of Lmax-Lmin and the determination of the image tone according to the calculation result is performed.
In this embodiment, the discrimination result “1” is output when it is determined that the pixel of interest is an edge portion such as a character, and the discrimination result “0” is output when it is determined that it is a halftone image. ing.

When the output of the ROM 13 is "1", "1" is input to the select terminal 5 of the multiplexer 14, and the A terminal input, that is, the pixel data of interest binarized by a certain threshold value is selected,
Output to 17. The printer 17 is a binary printer such as a laser beam printer, and forms dots according to the inputted binary data and records an image. On the other hand, when the output of the ROM 13 is "0", the B terminal input is selected and the dithered binarized signal is selected. In addition, delay RAM7,8 and D-
The F / F circuits 15 and 16 are used to match the output timings of the identification result and the target pixel.

With the configuration as described above, the image tone can be accurately identified for each pixel in real time and the image can be binarized. In this embodiment, the upper 2 bits of the 6-bit image data are directly input to the ROM as an address, and R
Since the determination result is output from the OM, the circuit scale can be small, and various algorithms can be applied without being limited to the above-mentioned determination method.

For example, when the pixel of interest has the maximum or minimum density value, even if the density difference ΔL is 2 or 3, it may be determined to be a halftone portion. By adding the positional relationship between the maximum and minimum density values in this manner, the image can be identified with higher accuracy.

As another embodiment, four pixels adjacent to the target pixel A may be set on a diagonal line as shown in FIGS. If the capacity of the ROM used is sufficiently large, eight adjacent pixels may be used for the determination as shown in FIGS.
Further, among the pixels shown in FIG. 3, the pixel of interest may be identified by using the pixel of interest A and the adjacent pixels B, D, J, and K. Further, the simple binarization circuit 3 and the dither processing circuit 4 may be constituted by a ROM or the like without using a comparator, and 6-bit image data may be directly input as an address and the processing result may be output.

〔effect〕

As described above, according to the present invention, in identifying the feature of the image represented by the image data of the pixel of interest, image data of a plurality of pixels adjacent to the pixel of interest is used, and an image of n bits (1 <n) is used. Since data is used, identification can be performed with high accuracy, and since n-bit image data, which is smaller than the number m of bits of input image data, is used, identification can be performed with a relatively simple configuration.

[Brief description of drawings]

FIG. 1 is a diagram for explaining an image identifying method in the present embodiment, FIG. 2 is a diagram showing an image processing apparatus in the present embodiment, and FIG. 3 is a diagram for explaining another identifying method. . Here, 1 is a document reading unit, 3 is a simple binarization circuit, 4 is a dither processing circuit, 5-8 are delay RAMs, 9-12, 15, 16 are delayed flip-flop circuits, 13 is a ROM, 14 is a multiplexer and 17 is a printer.

Claims (1)

[Claims] 1. An image processing apparatus for processing image data according to the characteristics of an image represented by the input image data, comprising: input means for inputting m-bit image data for each pixel; The input m-bit image data is converted into n bits (1 <n <m) including an intermediate level for each pixel, and the n-bit image data of the pixel of interest and the n-bit image of a plurality of pixels adjacent to the pixel of interest are converted. Identification means for identifying the feature of the image represented by the image data of the target pixel based on the image data; and m-bit image data of the target pixel input by the input means according to the identification result by the identification means. An image processing apparatus comprising: a processing unit for processing.