JPS6180963A - Picture signal processing method - Google Patents

Abstract

PURPOSE:To discriminate whether a notice picture element is a character image part or a halftone image part by deriving a density difference between the notice picture element and its peripheral picture element, multiplying these difference by N power and adding it to the peripheral picture element, and comparing a result of addition with a prescribed threshold level. CONSTITUTION:An A/D-converted digital signal is applied to an input terminal 11, and supplied to a series circuit of line memories 13, 17 and a series circuit of latches 141-143. Outputs of the line memories 13, 17 are supplied to series circuits of latches 144, 140 (notice picture element), and 145 and latches 146-148, respectively. Outputs of each latch 140-148 are supplied to a character half tone discriminating part 15, and in this part, a difference of density levels of the notice picture element (an output of the latch 140) and the peripheral picture elements (outputs of the latches 141-148) is derived. Subsequently, this difference is added to the peripheral picture element, and thereafter, compared with a prescribed threshold level, and a result of this comparison is supplied as a discriminating signal to a multiplexer 21. As a result, an output of a character processing part 22 or a halftone processing part 23 is selected and outputted.

Description

DETAILED DESCRIPTION OF THE INVENTION "Field of Industrial Application" The present invention is an image processing method for performing appropriate image processing on each image information source containing a mixture of character images and intermediate images. These pictures! (!FFI
The present invention relates to an image signal processing method for identifying I minutes.

``Prior art'' Image processing devices that read a document using an image sensor and record or display it using quantized image signals have a method of distinguishing between each part when the document contains text and photographs. Image processing is being performed. This is because pixel-by-pixel binarization processing (simple 2 (M ratio processing)) only reproduces text image parts well, making it impossible to reproduce halftone parts such as photographs, while dither processing etc. This is because while pseudo-halftone processing can satisfactorily reproduce halftones of photographs, etc., the reproduction of line drawings such as characters deteriorates.

As a conventional technique for distinguishing between a character image portion and a halftone image portion, a method disclosed in Japanese Patent Application Laid-Open No. 58-3374 is a typical one. In this method, the document surface (screen) is divided into blocks each consisting of multiple pixels.
The difference in density level is found between the pixel with the maximum density level and the pixel with the minimum density level within each block. If this difference is larger than a predetermined value, it is determined that it is a binary image area or a character image area, and otherwise, it is determined that it is a grayscale image area or a halftone image area.

"Problems to be Solved by the Invention" However, in such an image signal processing method, only the signal level of the pixel having the maximum or minimum value within the block is used as a basis for judgment, so it is difficult to determine whether the block contains noise. In this case, there was a risk of making the wrong decision.

In view of these circumstances, it is an object of the present invention to provide an image signal processing method that can stably discriminate between character images and halftone images from an image information source containing both characters.

"Means for Solving the Problem" In the present invention, differences in density levels between pixels are determined between a pixel of interest and surrounding pixels existing in a predetermined area surrounding the pixel, and these differences are calculated. The value raised to the Nth power is added to all surrounding pixels. The result of this addition is compared with a predetermined threshold value, and based on the comparison result, it is determined whether the pixel of interest belongs to the character image portion or to the intermediate or sub-pixel portion.

Here, N is a positive number, and may include a value below the decimal point. Typically, N=1 or N=2 can be considered. Note that the determination result obtained for one pixel of interest can also be used as the determination result for a group of pixels including surrounding pixels.

"Example" The present invention will be described in detail below with reference to Examples. ,.

'''11 11th ffllt This implementation PIT Image 0 judgment IC 1t゛63°31 pixel group P, ~P8 is represented. Of these, the central pixel P is the pixel of interest as the judgment target. and the eight pixels surrounding it, P,
~P8 are peripheral pixels. Each pixel Po=Pa
concentration d. -d8 is expressed as a signal level after A/D conversion of the read output of an image sensor (not shown). Concentration d. -d8 is expressed as 6-bit (671st floor) digital data.

In this embodiment, the pixel of interest is P. concentration d. and surrounding pixels P, -
The difference between the densities d and d8 of P is taken, squared and added, and the variance σ of the signal level in the 3×3 pixel group is investigated. Here, the variance σ is expressed by the following equation.

Then, the variance σ and the predetermined threshold j! Find the size relationship of Tll and select the pixel P of interest. It is determined whether the image belongs to the character image portion or the halftone image portion. This relationship is as shown in the following formula.

(i) σ>jeto ......Character image (ii
) σ≦β □ . . . The principle of discrimination for halftone images and above will be explained with reference to FIGS. 2 and 3. Figure 2 shows the portion of the manuscript consisting only of character images, 64
It shows the relationship between the signal level of the step (8 degree level) and the number of pixels of each level. Here, the signal level “′1
``signal level par 64'' refers to the case where that part of the document is ideally white, and the signal level 64 refers to the case where that part is ideally black. In a character image, there are two large peaks: signal level S in the ground color (background) part of the document, and signal level S2 in the so-called black part marked with ink, etc., and there are relatively large density differences even in small areas of the document. Usually there is. On the other hand, in halftone images such as photographs, a substantially uniform density distribution is obtained as shown in FIG. 3, and when a document is divided into small areas, there is little change in density within the area.

In other words, if the variance σ of the signal level is large in a relatively small area, that area has a strong tendency to be a character image, and conversely, if ikσ is small, that area has a strong tendency to be a halftone image (an elephant.Threshold ITH The optimum first shift is set by comprehensively considering the background color of the original, the density of the pencil and ink used, etc.

FIG. 4 shows an image signal processing circuit that performs the above-described discrimination processing. An input terminal 11 of this circuit receives a digital image signal 1 output from the A/D converter described above.
2 is manually input one pixel at a time. This digital image signal 12 is supplied to a line memory 13, and a series circuit consisting of first to third latches 14, 143, which latch the parallel 6-bit digital image signal 12 pixel by pixel. Three pixels are latched. These latch outputs are supplied to the character/halftone discrimination section 15 as signals representing the densities d1 to d3, respectively.

On the other hand, the digital image signal 12 supplied to the line memory 13 is delayed by a time corresponding to one latch of the image sensor that performs line sequential scanning, and is supplied to the line memory 17 as a digital image signal 16. Fourth, zeroth and fifth latches 144.14 having the same configuration as described above.
o, 145, and are sequentially delayed by one pixel and supplied. These lies/sochi 14. ．． The outputs of 14o and 145 are
The signals representing the densities d6, do, and ds are supplied to the 3+1-IJ separate section 15 for characters and intermediates, respectively.

The digital image signal 16 supplied to the line memory 17 is also read out after being delayed by a time corresponding to one line, and is read out by the sixth to eighth latches 146 to 146 having the same configuration as described above.
are sequentially delayed by one pixel and supplied. The outputs of these latches 146-148 have concentrations d6-d8, respectively.
is supplied to the character/halftone discriminator 15 as a signal representing the .

The character/halftone discrimination section 15 is composed of R, OM (!J-only memory), and a total of 54
A determination result signal 19 is output using the bit parallel signal as address information. This discrimination result signal 19 is obtained by calculating the fraction σ using equation (1) for each combination of latch outputs and using the threshold value A.
The results of comparison with TH are stored in advance.For example, in the case of a character image, the signal state will be H (high) level, and in the case of a halftone image, the signal state will be L (low) level. It has become. The determination result signal 19 is used as a control signal for the multiplexer 21.

The multiplexer 21 receives the target pixel P from the character processing section 22 and the halftone processing section 23, respectively. Processed image signal 24.2
5 are manually input in parallel, and the discrimination result signal 1
Either one of the processed image signals 24 or 25 is selected depending on the signal state of the signal 9, and is output as an image signal 26. Here, the character processing section 22 is a section that performs simple binarization processing, for example, and in the case of this embodiment,
Four levels of signal levels are quantized into two levels for each pixel of interest. Further, the halftone processing unit 23 individually sets threshold values using, for example, a dither method, and performs pseudo halftone processing.

This halftone processing section 23 may perform binarization processing on each dot constituting the dithered I-IJ hook, or may perform ternarization processing or more multi-value processing.

The image signal 26 obtained in this way is subjected to appropriate processing according to the image state to which the pixel of interest Pa belongs, and the image is well reproduced in the subsequent recording section or display section (not shown). Become.

In the example described above, a character image and a halftone image (elephant) were distinguished using a pixel area with a 3 x 3 matrix configuration, but it is free to use a pixel area with a 3 x 3 matrix configuration. In this embodiment, the case where the power iN is 2 has been described, but N can be any positive number (including non-integers). Depending on the image signal processing, N = 1, and individual It is also effective to perform discrimination by comparing the sum of absolute values of density differences with a threshold value.

Furthermore, in the embodiment, it was decided that all pixels (character image or halftone image) were determined one by one in order of roughness.
It is also effective to extend and apply the determination result of a specific pixel of interest as the determination result of a group of images in a predetermined range including the pixel of interest from the viewpoint of simplifying signal processing.

"Effects of the Invention" As explained above, according to the present invention, character images and halftone images are discriminated using all density information about the pixel of interest and pixels in a predetermined range located around it, so the discrimination results are reliable. Not only is this method highly accurate, but also the pixel to be processed, that is, the pixel of interest, is always located at the center of the determination area, making it possible to perform highly accurate determination. Further, according to the present invention, there is no need to compare density differences among all pixels in a block or pixel group, so the apparatus for image signal processing can be simplified.

[Brief explanation of drawings]

The drawings are for explaining one embodiment of the present invention.
The figure is a layout diagram showing the arrangement of the pixel of interest and surrounding pixels.
The figure is a characteristic diagram showing an example of density distribution for a character image.
The figure is a characteristic diagram showing an example of density distribution for a halftone image, and FIG. 4 is a block diagram of a signal processing circuit that implements the image signal processing method of this embodiment. 13.17...Line memory, 14...
・Latch, 15...Character/halftone discrimination section, P,
...Pixel of interest, Pl to P8 ...
Surrounding pixels. Applicant: Fuji Xerox Co., Ltd. Representative
Person Patent Attorney Umeo Yamauchi Figure 1 Figure 2 Figure 3 ′

Claims (1)

[Claims] 1. In an image processing device that performs different image processing on a character image portion and a halftone image portion other than this, a pixel of interest as a pixel to be subjected to image processing and a pixel of interest surrounding this pixel of interest. Find the difference in the density level between each of the surrounding pixels located in the peripheral part of the predetermined range, multiply these differences by a positive number N, add up all the values, and add this addition result to the predetermined value. An image signal processing method characterized by comparing the pixel with a threshold value and determining from the comparison result whether the pixel of interest belongs to a character image portion or a halftone image portion. 2. The image signal processing method according to claim 1, wherein the positive number N is 1. 3. The image signal processing method according to claim 1, wherein the positive number N is 2. 4. The image signal processing method according to claim 1, wherein the determination result obtained for the pixel of interest is used as the determination result for a group of pixels in a predetermined range including the pixel of interest.