JPS6135680A - Picture identification method - Google Patents

Abstract

PURPOSE:To discriminate whether an identification unit area is a dot gradation picture or not by obtaining a variance of a black picture element number constituting a black picture element area in the identification unit or the variance in the distance between the black picture element areas through the binary-coding of a picture. CONSTITUTION:Figures A, B show an example binary-coding a dot gradation or a continuous gradation picture. In case of the dot gradation picture, the variance of the black picture element numbers constituting each black picture element area and the variation of the distance between the black picture element areas are small, but in case of the continuous gradation picture, the variances are large. In case of a line drawing (not shown), the variances are large. Whether or not the unit area is the dot gradation picture is discriminated by obtaining the variance (standard deviation ) of the black picture element constituting each black picture area at each identification unit and the variance of the distances between black picture element area. (standard deviation)

Description

DETAILED DESCRIPTION OF THE INVENTION (Field of Industrial Application) The present invention relates to an image identification method for classifying and identifying an image into each area of a line drawing, a halftone image, and a continuous tone image.

(Prior Art) As a conventional image identification method, for example, the method shown in the following document is known.

■Mitsuhiro Muratsumi, Toshifumi Sakai, “Extraction of structural information in document images” Showa 5th National Conference 7 H-1 ■Morisumi Kurose, Akira Ejiri-9 Sakura j1 “Real time of halftone images in document images” Identification" 1981, Information Processing Society of Japan 23rd National Conference 6 C-6 ■Hiroshi Makino, Red 1■Shaku" Binary Reproduction of Document Images Containing Dark and Light Areas, March 1980, Theory of IEICE (D ) J65-D. 3.
IIEl. 307-314 ■Ken Takahashi-1 Masamitsu Ota “Computer-based image classification for compression coding” August 1980, IEICE Theory (D) J65-D.8.111), 101
8-102 ■ JP-A No. 56-132061 ``Halt image Shintan detection method'' ■ Yutaka Ueno, Takayoshi Semasa "Tone reproduction in newspaper halftones" Print magazine, 1982 (Vol. 66) 1, 01 )
.

15-23 ■ JP-A-56-1 No. 49674 ``Method for identifying image characteristics'' (Problems to be solved by the invention) These conventional methods are used to identify line drawings and continuous tone drawings, or to distinguish between line drawings and halftone dots. The main focus was on identifying whether it was a gradation painting or not.By the way,
In reality, most images are a mixture of line drawings, halftone gradation drawings, and continuous gradation drawings, and a method for identifying the above three types is also required. From this point of view, the present inventor presented the 116th National Conference of the Institute of Electronics and Communication Engineers in 1988, ``Study of Image Identification Methods.''
We proposed a method to distinguish the above three types using η. but,
This method has the problem that it is easy to misidentify continuous tone images with rapid density changes.

The present invention has been made in view of this problem, and its purpose is to provide an image identification method that can distinguish a continuous tone image from a continuous tone image regardless of whether the image has a rapid change in density or not.

(Means for Solving the Problems) The present invention solves the above problems by binarizing an image and detecting variations in the number of black pixels constituting the black pixel area within the identification unit area or the area of the black pixel area. At least one of the variations in the distance between them is determined, and based on the variation, it is determined whether or not the identified unit area is a halftone gradation image, and the identified unit area is determined from the proportion of the background. It is characterized by determining whether it is a line drawing or not.

(Example) Hereinafter, an example of the method of the present invention will be specifically explained using the drawings.゛In the method described here, first, an image is divided into multiple identification unit areas (
Line drawing, halftone gradation drawing, continuous lit! (minimum unit for determining which one is one). For example, if the image is NxM=2000x2500 pixels (1
0 dot /+u), the size of the identification unit area is set to k x k = 32×32 pixels. Next, the image is divided into two numbers, and a black pixel area is determined for each pixel for each identification unit area. Here, one black pixel area is, for example, a set of black pixels that are adjacent to each other in at least one of the four directions (up, down, left, and right), and there is no black pixel adjacent to one black pixel. In this case, a single black pixel forms one black pixel area. Note that the black pixel referred to here refers to a white pixel in a black-and-white inverted image. Figure 2 (a),
(mouth) has 1 halftone image and 2 continuous tone images, respectively.
An example of a value conversion case is shown (parallel vertical lines (pair) in the figure indicate black pixels). It can be seen that in the case of a halftone image, the variation in the number of black pixels constituting each black pixel area and the variation in the distance between black pixel areas is small, but in the case of a continuous gradation image, these variations are large. Although not shown, there are large variations in line drawings as well. Focusing on this point, the method of the present invention calculates the variation (standard deviation, etc.) in the number of black pixels constituting each black pixel area and the variation (standard deviation, etc.) in the distance between black pixel areas for each identification unit area. , it is determined whether the image is a halftone gradation image or not. Incidentally, FIG. 3 shows the black pixel area A1. A2
This shows an example of the distance measurement position when calculating the distance between
It shows that the distance between each black pixel area 1 and the center of gravity of A2 is measured. FIG. 4 shows the standard deviation of the number of black pixels within the black pixel area and the standard deviation of the distance within the black pixel area obtained in this way, with the former taken as the vertical axis and the latter as the horizontal axis. In the figure, the black circles are those obtained for some of the discrimination unit areas of the halftone gradation image, and the viewing angles are those obtained for some of the other discrimination unit areas. In the case of the halftone dot 111i1i image, it can be seen that both the standard deviations of the number of black pixels and the distance between black pixel areas are smaller than the others, and this tendency remains the same even when many samples are used. Therefore, in the method of the present invention, whether or not the corresponding identification unit area is a halftone gradation image is determined based on whether both or one of these standard deviations (especially the standard deviation of the number of black pixels) is smaller than a predetermined value. . For example, if the standard deviation of the number of black pixels is within 1.0 pixels, it is determined that the image is a halftone gradation image.

Next, in the method of the present invention, the proportion of dorsals is examined. This percentage is
Since the size of the identification unit area is fixed, it is proportional to the number of white pixels within the identification unit area, so it can be easily calculated by finding the total number of white pixels or the total number of black pixels within the identification unit area. However, without actually calculating it, the total number of white pixels or the total number of black pixels may be used as data indicating the ratio of background pixels. FIG. 5 shows the proportion of this background, and the horizontal axis shows each sample in the medium discrimination range.

In the figure, the black circles are for identification unit areas of continuous tone drawings, and the viewing angles are for line drawings. In the case of line drawings, since the proportion of the background is large, the difference is large compared to continuous tone drawings. Therefore, it is easy to determine whether an image other than a halftone gradation image is a line drawing or not.

By the above method, line drawings, halftone gradation drawings, and continuous gradation drawings can be identified.

By the way, the above explanation is for the case where the discrimination is performed over the entire area of the image, but if the image area is extracted, the above method may be applied to each image area. In this way, the area range to be identified becomes narrower, so that efficient identification can be performed.

Therefore, an example of a method for extracting an image area will be described below. FIG. 6 is a flowchart showing the procedure for t- to a member example of this extraction method. In this example, first, the background level (
density level), then block the image,
The pixels in each block are binarized using the same value determined based on the background level, and the texture analysis of each block is performed from the binarized image data to classify each block into a predetermined pattern. Check the connection of
Similar image regions are extracted. Note that the order of background level determination and image blocking may be interchanged or changed.

Next, the above procedure will be explained in detail step by step.

(1) Determination of the background level In order to determine the background level in a short time, a window of mxm (for example, 32x32) pixels as shown in FIG. 7 is provided on the image area, and the image within the window is Based on the data, a density histogram as shown in Fig. 8 is created and the background level is determined. Fig. 9 is a flowchart showing the procedure for determining the background level. First, a window is determined (position determined), and based on the window Create a IIa histogram and calculate the average value M and standard deviation S.

Next, if the standard deviation S is smaller than a predetermined threshold T, the background level B is selected as a value within the range of M-38≦B≦M+38. Usually, B=M-1-38. On the other hand, if the standard deviation S is below the threshold value or more, the window position is moved in the horizontal or vertical direction in FIG. 7, a new window is determined, and the density histogram is created again. This is because in the case of S≧King, there is no window in a position where the background can be captured well. Note that the size below the threshold may be any value as long as it is within a range that can capture the background, but it is preferably a size that allows the window to be determined a minimum number of times.

(Ir) Blocking the image Divide the image into multiple blocks. For example, if the image is NXM = 2000 x 2500 pixels (10 dots) as shown in Figure 10,
/mm), then the size of each block is n x n
Take =32X32. Of course, 16X16 or 64X64
You can also choose.

(I [[) Texture analysis of each block Next, the image data is binarized using the pre-coating phase level B as a threshold, and this results in a gain! , = Classify each block into a predetermined pattern using binarized image data 3. As a pattern to be prepared in advance, for example, tIlIl and 1f shown in FIG.
: Use 5 pieces (0 to G). This pattern changes from O to G, i.e. (0,1)-10(2-5)→(
6,・7) → (8~B) → (C~F) → G The pattern has many black and white changes in the season. The classification of each block into L patterns is determined by projecting the binarized image data in two-dimensional directions (XY directions) and calculating the cumulative number of black pixels ("1"), as shown in FIG. 12, for example. Pattern classification can be easily performed using a computer or the like by using a method of determining the corresponding pattern from the two integrated number distributions in the X and Y directions. In the example shown in FIG. 12, black pixels are concentrated at the bottom right of the block (), so the cumulative number on the right side is large in the X direction, and the cumulative number on the bottom side is large in the Y direction, and pattern 8 Since the same tendency is shown when projected in two-dimensional direction, it is concluded that the block corresponds to pattern 8.

(TV > Image area extraction Connections between each block may be examined on a block-by-block basis, or each block may be divided, for example, into two vertically and horizontally (a total of 4
Figure 13 is a simplified explanatory diagram to show the latter method, and Figure 13 (a) shows the image divided into blocks and a binarized image superimposed on it. (The shaded area is the black pixel area), and Figure 13 (b) shows the first pixel area in each block.
Figure 13 (c), which is a diagram with pattern numbers in Figure 1, shows the result obtained by dividing each block into four (for example, block α→small block α, α2.α3.α4) as described above. 12 is a diagram in which each small block is given the pattern number of FIG. 11. FIG.

Divide each pattern in Figure 11 into four.1. :, it is possible to know in advance which patterns each pattern is decomposed into, so it is extremely easy to obtain FIG. 13(c) from FIG. 13(b). Furthermore, when each pattern in Fig. 11 is divided into four, the 11th
Although each pattern shown in the figure may be selected, it is sufficient that the pattern can be approximately decomposed into the patterns shown in FIG.

Next, the connection between the small blocks is examined, and in order to judge that there is a connection relationship, for example, the criteria for determining whether there is a connection relationship is that there is at least a black part on the joining edge of patterns of adjacent small blocks. Alternatively, the presence or absence of a connection relationship may be determined by taking into consideration the overall shape of the pattern. In any case, the combinations of joining patterns are limited, so if you decide in what positional relationship each pattern will be adjacent to each other, there will be a connection relationship between small blocks.
Determining whether or not small blocks are connected is extremely simple. Figure 13 (d) is a diagram showing an image area obtained by assuming that there is no connection relationship between small blocks when there is only a white part on the joint side of the pattern. 4 & area. Of course, the connections between blocks may be checked using the above method without performing the four-division process. However, in this case, the boundaries of the regions become rough.

In order to solve this problem of roughness, it is conceivable to reduce the size of the block from the beginning, but doing so would require a large amount of time for the texture analysis.

As a result of the above, the image area has been extracted.

Figures 14 to 16 are more specific illustrations of region extraction, with Figure 14 showing the original image, Figure 15 showing the pattern classification image, and Figure 16 showing the region extraction image. It is. Note that the areas indicated by 1 and 2 in FIG. 14 are filled with characters, and the areas indicated by 3 and 4 are continuous tone images.

As shown in Fig. 15, the image extraction is successful, but in Fig. 15, isolated areas (blocks) occur due to dirt etc. in the original image, and these form unique image areas in Fig. 16. There is. In order to prevent this noise, an isolated block may be treated as having the same Ili condition as surrounding blocks.

After extracting image regions as described above, by applying the above-described identification method to each extracted image region, efficient image identification can be performed.

(Effects of the Invention) As explained above, in the method of the present invention, halftone gradation images are determined based on variations in the number of black pixels, line drawings are determined based on the proportion of the background, and the remaining gradation images are determined as continuous gradation images. Therefore, continuous tone images with rapid density changes will not be erroneously identified.

[Brief explanation of drawings]

Fig. 1 is an explanatory diagram showing an example of a discrimination unit area, Fig. 2 is an explanatory diagram of a binarized image, Fig. 3 is an explanatory diagram of distance measurement between black pixel areas, and Fig. 4 is an explanatory diagram of a distance measurement between black pixel areas. Figure 5 showing the relationship between the standard deviation of distance and the standard deviation of the number of black pixels constituting a black pixel area.
The figure shows the ratio of the background, Figure 6 is a flowchart showing the procedure for implementing an example of the area extraction method, Figure 7 is an explanatory diagram of a window, and Figure 8 is an explanatory diagram of density hiss [Durham]. Fig. 9 is a flowchart showing the background level determination procedure, Fig. 10 is an explanatory diagram of image blocking, Fig. 11 is an explanatory diagram of block patterns, Fig. 12 is an explanatory diagram of pattern classification, and Fig. 13 is an explanatory diagram of the block pattern. FIG. 14 is a diagram showing an original image, FIG. 15 is a diagram showing a pattern classification image, and FIG. 16 is a diagram showing a region extraction image. α...Block α1. α2. α3. α4...Small block 1.2...
Character drawing area 3.4... Continuous tone drawing area Patent applicant
Roku Konishi Photo Industry Co., Ltd. Representative Patent Attorney Fuji Ijima 1 name bird 1 figure bird 3 figure falcon 4 figure yIh pixel area Wf) NU) 85414 difference CI) ixe
l) Horse 7 figure Horse 8 figure concave Concentration level Jv19 figure 1o figure

Claims (5)

[Claims] (1) Binarize the image to find the variation in at least one of the variations in the number of black pixels constituting the black pixel area in the identification unit area or the variation in the distance between the black pixel areas, and from the variation An image identification method that determines whether or not the identification unit area is a halftone gradation image, and also determines whether or not the identification unit area is a line drawing based on the proportion of the background. (2) The image identification method according to claim 1, wherein an image divided into a plurality of parts is used as the identification unit area. (3) Divide the image into multiple blocks, classify each block into a predetermined pattern using binarized image data, check the connection between each block using the pattern, extract the image area, and extract the extracted image. 2. The image identification method according to claim 1, wherein an area is used as the identification unit area. (4) The image identification method according to claim 3, characterized in that when classifying the blocks into patterns, data obtained by two-dimensionally projecting the binarized image data of each block is used. (5) When checking the connections between the blocks, each of the blocks is divided into two vertically and horizontally, and the connections between the divided small blocks and the surrounding small blocks are checked. The image identification method described in item 1 or 4.