JPH0457274B2 - - Google Patents

Description

Detailed Description of the Invention "Field of Industrial Application" The present invention also applies to image information sources in which three types of images, character images, photographic images, and halftone dot images, coexist. The present invention relates to an image signal processing device for performing.

``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 binarization processing) only reproduces text image parts well, making it impossible to reproduce halftone parts of photographs, etc., while pseudo-binarization processing such as dither processing This is because while halftone processing can satisfactorily reproduce the halftones of photographs, etc., the reproducibility of line drawings such as characters deteriorates.

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

By the way, some manuscripts use halftone dots to reproduce halftones or emphasize specific parts of text. The manuscript portion using such halftone dots (halftone image) is
If the image is read as is and subjected to image processing, moiré will occur due to the relationship between the halftone dot repetition period and the sub-scanning period, and the reproducibility of the image will be extremely degraded. Therefore, conventional techniques have been used to detect the halftone image portion existing on the document and to change image processing between this portion and other portions. A typical halftone dot detection device for this purpose is the technique described in Japanese Patent Application Laid-Open No. 58-218271. This device compares a binary image signal obtained by scanning a document with a detection signal consisting of a black and white code string corresponding to the positions and sizes of white and black dots on the document. By determining whether these signals match, the halftone image area is distinguished from other non-halftone image areas.

In this way, conventional techniques have been used to distinguish between character image parts and halftone image parts, and halftone image parts and non-halftone image parts. Helpful. However, a document contains a mixture of three areas: a character image, a halftone image, and a photograph. Here, the character image portion is a portion composed of line drawings such as characters. In addition, the halftone image part is the part where halftones are expressed, and is the part of the image where the gradation is expressed by halftone printing, and the so-called "halftone" part in the printer.
refers to the part of the image where the process is being performed. Furthermore, the photographic portion is a portion in which halftones are similarly expressed, and refers to image portions other than character images and halftone dot image portions. Photographs using a photosensitive substance such as silver chloride can be cited as a representative example of this photographic part.

``Problems to be Solved by the Invention'' Now, with regard to manuscripts in which three types of image parts coexist: a character image part, a halftone image part, and a photograph part, it has traditionally been said that all three types of image parts are good. There was no way to reproduce it. That is, in an image signal processing device that processes character images and halftone images separately, moiré occurs in halftone dot areas. In order to prevent the occurrence of moire, the halftone image portion is smoothed and then pseudo halftone processing is performed using a dither method or the like. However, when such smoothing is performed, a new problem arises in that the sharpness of the photographic part decreases.

On the other hand, in the image signal processing method that processes halftone images and non-halftone images separately, since text and photographic parts are included in the non-halftone areas, either the reproduction of text or the reproduction of halftones is affected. There was a problem in that if one was attempted to be performed satisfactorily, the reproducibility of the other would be insufficient.

"Means for Solving Problems" In the present invention, as shown in principle in FIG. It is made to sequentially determine which part of the processing object belongs to the entire area of the processing target. and image discrimination section 11
The image processing unit 12
Perform optimal processing within the system.

Part 1 that processes character images in the image processing unit 12
4 includes an MIF correction means 15 and a multi-value conversion means 16 for converting the image signal corrected by the MIF correction means 15 into multi-value. The halftone dot image processing section 17 in the image processing unit 12 includes halftone dot removal means 18 that smoothes and removes halftone dots, and halftone dot removal means 18 that performs pseudo halftone processing on the image signal after halftone dot removal. processing means 19. Furthermore, the portion 21 that processes photographic images in the image processing includes a halftone processing means 22 that performs multi-value pseudo halftone processing.

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

FIG. 2 shows an image signal processing device used in a document reading device as an embodiment of the present invention. An image sensor 31 of this device raster-scans an original (not shown) and generates an analog image signal 32.
Output. The A/D converter 3 quantifies this pixel by pixel and outputs it as a 6-bit (64 levels of density) parallel digital image signal 34. The digital image signal 34 is supplied to a determination circuit 35 to determine the image state, and is also supplied to a halftone image processing section 36, a photographic image processing section 37, and a character image processing section 38 for image processing. It's summery.

In the determination circuit 35, for each predetermined signal processing area,
It is determined whether the image belongs to a character image, a photographic image, or a halftone dot image. In this embodiment, such a determination is made by analyzing the constituent frequency components of an image within a window called a signal processing area. For example, FIGS. 3 and 4 show
The same image is expressed by a photograph and halftone dots, respectively, and is expressed for one scanning line. As can be understood from these figures, both exhibit smooth density changes when viewed macroscopically. However, when viewed microscopically, a photographic image still exhibits smooth density changes, whereas a halftone image reproduces the black and white dots that make up the halftone dots, causing considerable vibrational variation in density. .

Therefore, if Fourier transform is performed for each predetermined signal processing region as frequency analysis, in the case of a photographic image, a peak will occur only in a low frequency portion corresponding to the overall density change. On the other hand, in the case of a halftone image, in addition to one peak occurring in this low frequency portion, a peak also occurs in a higher frequency portion corresponding to the vibration portion (halftone dot). FIGS. 5 to 7 show typical examples of Fourier transform outputs after quaternary conversion for halftone dots, photographs, and character images, respectively.

FIG. 8 specifically shows a determination circuit that performs determination based on the principle described above. The digital image signal 34 is input to a shift register 41 of this circuit. The shift register 41 has a 33-stage configuration, and the pixel to be subjected to image processing (pixel of interest)
, and accumulates image signals for 16 pixels on each of the left and right sides. The area of the image signals stored in this shift register 41 corresponds to the signal processing area in the present invention.

The image signals of this signal processing area are inputted in parallel to the Fourier transform circuit 42 and subjected to Fourier transform. The converted output is input to the tetrachloride circuit 43 for each spatial frequency component and is converted into four values. The 2-bit 4-valued data outputted from each 4-valued circuit 43 is supplied to a determination ROM (read-only memory) 44 as address information. For judgment
The pattern data shown in FIGS. 5 to 7 is input to the ROM 44, and one of three types of images is read out as a judgment result according to the contents of the address information, and output as a judgment result signal 45. . The determination result signal 45 is supplied as a 2-bit signal to a multiplexer 47 shown in FIG.

Now, in the image signal processing device shown in Fig. 2,
A halftone image processing section 36, a photographic image processing section 37, and a character image processing section 38 each process the digital image signal 34 in parallel. In the halftone image processing section 36, the digital image signal 34 is first smoothed in a halftone removal circuit 48.
Small variations in shading characteristic of halftone images are removed. The image signal 49 after halftone removal is input to a pseudo halftone processing circuit 51. The pseudo halftone processing circuit 51 sets a threshold value for each pixel using, for example, a dither method, and performs pseudo halftone processing. At this time, each dot constituting the dither matrix may be expressed in two values of white and black, or may be subjected to three-value or more multi-value processing. The image signal processed by the pseudo intermediate processing circuit 51 is delayed by a predetermined amount in a delay circuit (not shown), and is supplied to the input terminal of the multiplexer 47 as the image signal 52 of the pixel of interest.

Next, the photographic image processing section 37 directly inputs the digital image signal 34 to the pseudo halftone processing circuit 53,
The same processing as the pseudo halftone processing circuit 51 is performed. The processed image signal is similarly delayed by a delay circuit (not shown) and is supplied to the other input terminal of the multiplexer 47 as the image signal 54 of the pixel of interest.

Finally, in the character image processing section 38, the digital image signal 34 is input to the MTF correction circuit 55, where deterioration in resolution caused by the optical system is corrected. The corrected image signal 56 is sent to the multilevel conversion circuit 5
7, where it is multivalued pixel by pixel.
The multivalued image signal is delayed by a delay circuit (not shown) and is supplied to the remaining input terminals of the multiplexer 47 as the image signal 58 of the pixel of interest.

The multiplexer 47 selects one of the three types of image signals inputted in parallel in this way according to the content of the determination result signal 45, and selects the image signal 59.
and output. That is, for example, when the determination circuit 35 determines that the pixel of interest is a pixel of a halftone image, it selects the image signal 52 that has been processed for the halftone image and outputs it as the image signal 59. Similarly, when the pixel of interest is determined to be part of a photographic image, the image signal 54 is selected, and when the pixel of interest is determined to be part of a character image, the image signal 58 is selected.
As a result, the image signal 59 output from the multiplexer 47 is always subjected to optimal image processing, and the image is reproduced satisfactorily in a recording section or display section (not shown) that uses this image signal 59. becomes.

Note that in the embodiment described above, the image state is determined for all pixels one by one, but the determination result of a specific pixel of interest is extended and applied as the determination result of a group of pixels in a predetermined range including this pixel of interest. This is also effective from the viewpoint of simplifying signal processing and removing noise. Further, in the embodiment, a pseudo halftone processing circuit is provided for each of the two processing sections, but a switching circuit or the like may be provided so that one pseudo halftone processing circuit is shared by these processing sections. Further, in the embodiment, digital processing is performed using a pseudo-halftone processing circuit, but it is also possible to express halftones using an area gradation method that allocates patterns with printing ratios corresponding to gradations.

"Effects of the Invention" As described above, according to the present invention, Fourier transform is performed for each predetermined signal processing area that is the target of image processing on the same scanning line, and the resulting spatial frequency output distribution is If it contains a high frequency, this signal processing area is determined to be a character area, and in other cases, if the peak position exists only in a low frequency part, it is determined to be a photographic area. Further, if there are peaks in the lower frequency portion and the higher frequency portion, it is determined that the area is a halftone dot area.
Therefore, it is not only possible to distinguish between these three regions on a scanning line basis, but also on the same scanning line. Even when text areas, photo areas, etc. are complex, these image signals can be processed satisfactorily. Furthermore, since these three types of discrimination can be performed simultaneously, if the information for these discriminations is written in, for example, a read-only memory, the discrimination results can be immediately obtained for each signal processing area with a simple circuit configuration. can. In addition, since optimal signal processing is performed for each, it is possible to ensure sufficient sharpness for character images and photographic images, and also to perform good reproduction of halftone dot images.

[Brief explanation of the drawing]

Fig. 1 is an explanatory diagram for explaining the principle and deformability of the present invention, Fig. 2 is a block diagram of an image signal processing device according to an embodiment of the present invention, and Fig. 3 is a diagram of reflectance characteristics of a photographic image. Figure 4 is a characteristic diagram showing an example of reflectance characteristics of a halftone image, and Figures 5 to 7 are characteristic diagrams of spatial frequency characteristics in each image area of halftone dots, photographs, and characters. FIG. 8 is a block diagram showing a specific example of the determination circuit shown in FIG. 2, which is a characteristic diagram showing a typical example. 11... Image discrimination section, 12... Image processing section, 1
5...MTF correction means, 16...Multi-value conversion means, 1
8... Halftone removal means, 19, 22... Halftone processing means, 35... Judgment circuit, 36... Halftone image processing section, 37... Photographic image processing section, 38... Character image processing section, 48 ...Dot removal circuit, 51, 53...
Pseudo halftone processing circuit, 53...MTF correction circuit,
57...Multi-value circuit.

Claims (1)

[Scope of Claims] 1. Fourier transform means that performs Fourier transform for each predetermined signal processing area that is the target of image processing on the same scanning line; If the signal processing area contains a high frequency, it is determined that the signal processing area is a character area, and in other cases, if the peak position exists only in a low frequency area, it is determined that the signal processing area is a photographic area. , an image discriminating unit that determines that the area is a halftone dot area when there are peaks in both the low frequency part and the high frequency area; For the image of the region
After performing MTF correction, a character image processing unit converts the image into a multivalued image, and if the image discrimination unit determines that the image is a photographic area, performs multivalued pseudo-halftone processing on the image of the signal area. a photographic image processing unit that performs multi-value pseudo-halftone processing on the image from which halftone dots of the image in the signal area are smoothed and removed when the image discrimination unit determines that the image is a halftone dot area; What is claimed is: 1. An image signal processing device comprising: a halftone image processing unit that performs the above operations.