JPH067663B2 - Halftone image processing method - Google Patents

Description

TECHNICAL FIELD The present invention relates to a halftone image processing method in an image recording apparatus.

2. Description of the Related Art Most of hard copy devices such as facsimile receivers and dot printers are binary recording, and various systems have been proposed in which halftones are artificially expressed by binary dots of black and white. In the old days, there is a halftone dot method used for printing plate making, and recently, the dither method is widely used.

FIG. 3 (a) is a principle diagram of the well-known dither method. The grayscale pixel input signal x _ij at the coordinates (i, j) is compared with a certain threshold signal d _ij, and x _ij is successively thresholded to be converted into a binary dot signal g _ij . When d _ij has a constant value, it is a simple binarization process and halftone cannot be expressed.
When d _ij is changed stochastically with (i, j), the ratio of the output dots g _ij being turned on or off changes corresponding to the gray level of x _ij , so that the pseudo intermediate by the integration action of the eye. The key is perceived. A method of giving d _ij by random numbers irrespective of (i, j) is called a random dither method, and a method of arranging according to a certain rule spatially with respect to (i, j) is called an ordered dither method. b) shows a typical Bayer type systematic dither method. In this example, assuming that (i, j) is defined as a matrix of 4 × 4, and is projected on a display device such as a plasma display panel,
When the pixel level of the input image is higher than the threshold value, it is turned on (white), and when the pixel level of the input image is lower than the threshold value, it is turned off (black), and x _ij is shown to be large in the bright (white) part. There is. When x _ij is expressed by the density in the _printing apparatus, the black-and-white relationship is reversed, but this is convenient and any one may be used as long as the input-output density relationship corresponds.

Now, the quality of the image reproduced by the systematic dither method depends on the arrangement of the threshold matrix. Generally, when the dots are dispersed and arranged, the resolution is excellent, but the gradation is sacrificed. Conversely, when the dots are arranged in a concentrated manner, the dots are excellent, but the resolution is sacrificed. FIG. 4 shows an example of a typical 4 × 4 threshold matrix.
Shows a Bayer type (dispersion type) with good resolution, and (b) shows a spiral type (concentration type) with good gradation. Further, as an example of giving an intermediate property between (a) and (b), there is a halftone dot type of (c), etc., and the above is relatively often used.

Problems to be Solved by the Invention However, in the systematic dither method, the resolution and the gradation are contradictory as described above.

For example, in some cases, such as an image of a person, an image of an article, or an image of a landscape, depending on the image, it is necessary to emphasize the gradation, or conversely, to emphasize the resolution. .

However, it is difficult to obtain a proper image quality according to the target image by the conventional systematic dither method.

In view of the above-mentioned drawbacks of the prior art, the present invention relates to a halftone image processing method capable of obtaining an appropriate image quality according to a target image quality.

Means for Solving the Problems Each pixel value of the original image is defined by mM rows × nN columns (where m, n,
M and N are integers, and k (where k = m × n and k ≧ 2) level regions when converted into a binary signal by comparison with a threshold matrix composed of mM ≧ 2, nN ≧ 2) For the pixel value level of the original image, the threshold matrix is decomposed into a multilayer matrix of k layers of M × N elements, and each of the k layers is divided according to the calculated frequency distribution. The above object is achieved by determining the elements of the matrix.

Function The present invention is to flexibly adapt the structure of the threshold matrix composed of mM × nN to the systematic dither method according to the property of the target original image.

Embodiment An embodiment of the present invention will be described below with reference to the drawings. Note that, in the following description, when 64 gradations are realized by an 8 × 8 matrix, it is decomposed into 4 layers of 4 × 4 dither matrix.

Now, FIG. 1 (a) shows a histogram of pixel values of the target original image. Here, when the pixel value level is divided into four (l = 1, ..., 4 where l is an area number), and the number of pixels having a pixel value in each divided area is counted,
It becomes as shown in Fig. 1 (b). If the region numbers are arranged from this graph in descending order of distribution number, l = 4, l = 1, l = 3, l
= 1. Of these, l = 4 and l = 1 are assigned to the resolution-oriented dither method, and l = 3 and l = 2 are assigned to the tone-oriented dither method. An 8 × 8 dither matrix constructed in accordance with this is shown in FIG. When k = 1, 4 (k is a layer number), pixel values 0 to 15 and 48 to 63 are gradation-oriented type, that is, dot concentrated type, and when k = 2 and 3, pixel values 16 to 47 are resolution-oriented type, that is, dot. Allocate to distributed type and construct 8x8 dither matrix by combining these layers.
It suffices to perform the processing according to the dither method shown in.

The halftone image formed by the above embodiment is a mixture of both good characteristics in resolution and gradation.

In the present embodiment, k = 4 and 1 are set to the resolution-oriented type,
= 2 and 3 are the gradation-oriented type, but k depending on the image
= 4 and 1 may be a gradation-oriented type, and k = 2 and 3 may be a resolution-oriented type. If necessary, various combination methods are conceivable, such as only 1 = 4 being a resolution-oriented type and k = 1, 2, 3 being a tone-oriented type.

As described above, according to the present invention, it is possible to alleviate the disadvantage that the resolution characteristic and the gradation characteristic are remarkably sacrificed in the conventional systematic dither method, and the image quality according to the characteristic of the target image is reduced. Can be controlled.

[Brief description of drawings]

FIG. 1 (a) is a histogram for the original image value of the target original image for explaining the halftone image processing method in one embodiment of the present invention, and FIG. 1 (b) is a distribution graph for the original image of FIG. FIG. 2 is a conceptual diagram showing the structure of an 8 × 8 threshold matrix based on the target original image of FIG. 1, and FIGS. 3 (a) and 3 (b) are conceptual diagrams of the binarization method for halftone images. FIG. 4 is a conceptual diagram of a conventional threshold matrix of typical systematic dither.

Claims (2)

[Claims] 1. A binary signal by comparing each pixel value of an original image with a threshold matrix consisting of mM rows × nN columns (where m, n, M and N are integers and mM ≧ 2, nN ≧ 2). When converting to, the frequency distribution of the pixel value level of the original pixel is obtained for k (where k = m × n, k ≧ 2) level regions, while the threshold matrix is calculated from M × N elements. K
A halftone image processing method of decomposing into a multi-layer matrix of layers and determining the elements of each threshold matrix of the k layers according to the obtained frequency distribution situation. 2. The l-th frequency distribution (where l is l <k
(Integer value of) of the original image in the area up to (k-1) th, while the original image in the area up to (k-1) th is represented by the dot-distributed systematic dither method. The halftone image processing method according to claim 1, wherein elements of a threshold matrix of layers are assigned.