JPH06195457A - Image processor - Google Patents

Abstract

PURPOSE:To make it possible to eliminate the edge emphasis in an unnecessary direction by a simple circuitry by determining the presence or absence of the edge emphasis and the direction based on a binary pattern. CONSTITUTION:Image data to be inputted in a memory 1 regards the output value from a CCD as plural bits serial image data after a digital conversion is performed. In an edge direction selection part 20, the presence or absence and the direction of an edge emphasis is detected by the pattern matching after a binary processing is performed. In a four direction Laplacian calculation part 9, the Laplacian of the total four directions of a horizontal direction, a vertical direction, two oblique directions is calculated. Next, in four-direction one direction Laplacian selection part 10, the Laplacian in the edge direction selected in an edge direction selection part 20 is selected from the output of the four-direction Laplacian calculation part 9 and is outputted as edge emphasis amount to an edge emphasis amount calculation part 11. In this edge emphasis amount calculation part 11, the inputted edge amount is added to the value of the image data of a picture element under consideration and the image data is outputted from a Pout.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image processing apparatus for executing a process of emphasizing an edge of image data.

[0002]

2. Description of the Related Art Some digital copying machines use a Laplacian filter to perform edge enhancement for performing MTF correction on image data read by an image reader or the like (for example, Japanese Patent Laid-Open No. 62-114377). Gazette). Furthermore, in order to perform more appropriate edge enhancement processing,
Various methods have been proposed. For example, Japanese Patent Laid-Open No. 3-15
In Japanese Patent No. 7060, the presence / absence and the direction of an edge are detected by multilevel restoration of a binary image and pattern matching. Also proposed is multi-valued restoration of binary image data, and a method of obtaining the difference from the number of black pixels existing in a predetermined matrix for edge enhancement processing in a plurality of directions. Furthermore, Japanese Patent Application Laid-Open No. 61-157165 proposes a method in which edge amounts in a plurality of directions are obtained and the sum thereof is adopted.

[0003]

However, a Laplacian is used for a pencil-written manuscript whose density varies depending on the writing pressure of the recording person, or a recycled paper or a pamphlet whose density of characters varies in the background. When the edge enhancement process using the filter is executed, the density variation is subjected to the edge enhancement process. As a result, the characters on the document are blurred and partially thickened.

FIG. 18 is a diagram showing an example of a typical Laplacian filter. The Laplacian filter is mainly used in an MTF correction unit in a digital copying machine. The image data of the original image read by the image sensor (CCD) lacks sharpness due to the wraparound of light from adjacent pixels. However, sharpness is an important point, especially for characters and line images. Therefore, in the MTF correction unit, a Laplacian filter is used, and the output value of this filter is added to the density value of the pixel of interest as the edge enhancement amount.

However, the output value of the Laplacian filter is not always an accurate edge enhancement amount. For example, there may be variations in the output values between the pixels of the CCD, or an edge other than the boundary between the character portion and the background portion of the document. The latter case is a problem in the present invention, an example of which is shown in FIGS. 19 (a) to 19 (c).

Each of the drawings shown in FIGS. 19A to 19C corresponds to FIG.
The edge enhancement processing is performed on the halftone image data in each predetermined matrix using the aplacian filter 1 shown in FIG. In addition, each figure shown in FIG.
The same result can be obtained even with the Laplacian filter 2 shown in FIG. 18B by using a document rotated once.

With the Laplacian filter 1 shown in FIG. 18A, as shown in FIG. 19A, if the density inside the line drawing is uniform, a desired result can be obtained. However, as shown in FIGS. 19B and 19C, when the density value of the pixel in the center of the line drawing is higher or lower than the density of the adjacent pixel, this difference in density value is emphasized in the edge enhancement processing. This means that the edge enhancement processing as a line drawing cannot be performed. Such originals are generated when there are variations in density due to changes in writing pressure, such as in pencil-written originals, variations in the density of white background areas such as recycled paper, and pamphlets. As described above, there may be a case where a character exists in the halftone image.

Particularly, when it is necessary to repeat the copying operation a plurality of times (n times), the image once subjected to the edge enhancement processing is further subjected to the edge enhancement processing or the binarization processing. As a result, a part of the line drawing is missing as shown in FIG. In practice, this will appear on the copy sheet as a smeared character image from the copy.

Therefore, it is an object of the present invention to provide an image processing apparatus capable of performing more appropriate edge enhancement processing.

[0010]

An image processing apparatus according to a first aspect of the present invention is an image processing apparatus for executing edge enhancement processing of image data of a document image read by an image reader, wherein a pixel of interest is placed at the center. By using a first determining means for determining a threshold value for performing binarization based on a value of image data of pixels forming the first matrix, and a threshold value determined by the first determining means. First binarizing means for binarizing the image data in the first matrix; judging means for judging the presence or absence of edge enhancement processing based on the binarized data by the first binarizing means; When the edge amount detecting means for detecting the edge amount in the matrix of 1 and the edge enhancing processing are determined by the determining means,
Edge enhancement means for performing edge enhancement processing based on the detection result by the edge amount detection means.

An image processing apparatus according to a second aspect of the present invention is an image processing apparatus that executes edge enhancement processing of image data of a read original image, and forms a first matrix having a pixel of interest at the center. For determining a threshold value for binarization based on the value of the image data of the pixel to be converted
Detecting means, first binarizing means for binarizing the image data in the first matrix using the threshold value decided by the first deciding means, and detecting the amount of edges in the first matrix A threshold value for binarization based on the value of the image data of the pixels that form the second matrix having a smaller size than the first matrix In the second matrix, and a second binarizing means for binarizing the image data of the pixels forming the second matrix using the threshold value by the second deciding means. Second edge amount detecting means for detecting the edge amount of, the judging means for judging whether or not the binarized data by the first binarizing means satisfies a predetermined condition, and the binarizing data is judged by the judging means. If it is determined that the predetermined condition is not satisfied, the first First determining the presence or absence of the edge enhancement processing based on the binary data by the binarizing means
When the determination unit and the first determination unit determine to execute the edge enhancement processing, the first edge enhancement processing for performing the edge enhancement processing on the target pixel of the first matrix based on the detection result by the first edge amount detection unit. And second judging means for judging the presence or absence of edge enhancement processing based on the binarized data by the second binarizing means when the binarized data is judged by the judging means to satisfy a predetermined condition. If the second determination means determines to execute the edge enhancement processing, the second edge enhancement means for performing the edge enhancement processing on the target pixel of the second matrix is determined based on the detection result by the second edge amount detection means. Prepare

Still further, an image processing apparatus according to a third aspect of the present invention is an image processing apparatus for executing edge enhancement processing of image data of a document image read by an image reader, the first image processing apparatus having a pixel of interest at the center. Determining means for determining a threshold value for binarizing based on the value of the image data of the pixels forming the matrix, and the threshold value determined by the determining means. Binarizing means for binarizing, and a judging means for judging the presence / absence of edge enhancement processing based on the binarized data in the second matrix smaller in size than the first matrix, and in the second matrix When the edge amount detecting means for detecting the edge amount of the edge detecting means and the determining means determine to execute the edge enhancing processing, the edge enhancing processing is performed based on the detection result by the edge amount detecting means. And a edge enhancement means for lines.

The image processing apparatus described in claim 4 is the image processing apparatus described in claim 1, claim 2, or claim 3, in which the judgment means stores a plurality of binarized patterns. A memory for performing comparison, and compares the binarized data by the binarizing means with the pattern stored in the memory,
Based on this result, it is determined whether or not the edge enhancement processing is performed.

An image processing apparatus according to a fifth aspect of the present invention is the image processing apparatus according to the first, second, third or fourth aspect, wherein the determination means further includes edge enhancement processing. The direction is judged and the edge amount detecting means
Detecting edge amounts in a plurality of directions, the edge emphasizing unit selects an edge amount in the direction determined by the determining unit from the edge amounts in the plurality of directions detected by the edge amount detecting unit, and executes the edge enhancing process. .

[0015]

According to the image processing apparatus of the present invention, the image data in the first matrix is binarized by using the threshold value decided by the first decision means, and based on the binarized data. Whether or not the edge enhancement processing is executed is determined by the determination means. When the determination unit determines to execute the edge enhancement process, the edge enhancement unit performs the edge enhancement process based on the edge amount detected by the edge amount detection unit.

According to another aspect of the image processing device of the present invention, the threshold value is determined by the first determining means based on the image data of the pixels forming the first matrix, and the first threshold value is determined using this threshold value. The binarization means for binarizing the
Threshold value is determined by the second determining means on the basis of the image data of the pixels forming the second matrix having a smaller size than the matrix, and the threshold value is used by the second binarizing means. To convert. Here, when the judging means judges that the binarized data of the first matrix satisfies the predetermined condition, the second judging means uses the binarized data of the second matrix to make the second matrix. It is determined whether or not the edge enhancement processing is performed on the target pixel of. When it is determined by the second determination means that the edge enhancement processing is to be executed, the edge enhancement is performed on the target pixel of the second matrix based on the value of the edge amount in the second matrix obtained by the second edge amount detection means. Execute the process. Further, when the judging means judges that the binarized data of the first matrix does not satisfy the predetermined condition,
The first judging means judges whether or not the edge emphasis processing is executed for the target pixel of the first matrix. First
When it is determined by the determination means that the edge enhancement processing is to be executed, the edge enhancement processing is performed on the target pixel of the first matrix based on the value of the edge amount in the first matrix obtained by the first edge amount detection means. Run.

The image processing apparatus according to claim 3 is 2
When obtaining the threshold value used in the binarization process, the first matrix is used and binarization is performed using the threshold value determined by the determination means. Next, when the determination means determines whether or not to perform the edge enhancement processing, the second matrix having a smaller size than the first matrix is used for the determination. When the determination unit determines to execute the edge enhancement process, the edge enhancement process is performed based on the detection result of the edge amount detection unit.

An image processing apparatus according to a fourth aspect is the image processing apparatus according to the first aspect, the second aspect, or the third aspect, wherein a memory for storing a plurality of binarized patterns is provided in the judging means. The binarized data provided by the binarization unit is compared with the pattern stored in the memory, and the determination unit determines whether or not the edge enhancement processing is performed based on the result. When the determination unit determines to execute the edge enhancement process, the edge enhancement process is performed based on the detection result of the edge amount detection unit.

The image processing apparatus according to claim 5 is 2
Based on the data binarized by the binarizing means, the judging means judges the execution and the direction of the edge emphasis processing, the edge amount detecting means detects the edge amounts in a plurality of directions, and the edge emphasizing means sets the edge amount. The edge amount in the edge enhancement direction determined by the determination unit is selected from the edge amounts in the plurality of directions detected by the detection unit, and the edge enhancement process is executed based on this value.

[0020]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An image processing apparatus according to the present invention will be described below in detail in the following order with reference to the accompanying drawings. (1) Outline description of edge enhancement processing (2) Configuration of image processing apparatus (3) Detailed description of each functional unit

(1) Outline Description of Edge Enhancement Process First, the outline of the edge enhancement process used in the present invention will be described with reference to FIGS. 1 (a) and 1 (b). The left ends of (a) and (b) are diagrams showing the density data of the pixels in the 3 × 3 matrix area, and the density values of the pixels in the central vertical row are “light, dark, light” and “dark, light”. It is halftone data of "Dark". The halftone data is first binarized with a predetermined threshold value and then subjected to edge enhancement processing. In this embodiment, the average value of the density values of each pixel in the predetermined matrix is used as the threshold value. Unlike the case of executing the edge enhancement processing before performing the binarization as in the conventional example shown in FIG. 19, by executing the binarization processing first, (a),
As shown in the center of (b), even if the image density of the line drawing varies to some extent, all the pixels are determined to be black pixels. in this way,
Even for a line drawing such as a character image having a variation in image density, an appropriate edge enhancement process can be executed once the binarization process has been executed.

The binarized image data is then compared with a plurality of binarized patterns stored in the memory, and edge enhancement processing is executed only when the corresponding pattern exists. By doing so, it is possible to accurately check whether or not the edge enhancement processing needs to be executed, and it is possible to accurately determine the direction of the edge enhancement processing. This results in FIGS. 1 (a) and 1 (b).
In both images, it is determined that edge enhancement should be performed using the one-dimensional Laplacian filter in the horizontal direction (rightmost figure). Further, if the image density variation in the line drawing portion of the image data of the original image is sufficiently smaller than the edge emphasis amount, the line image density variation in the image after the edge emphasis is not so noticeable.

Here, after binarizing the image data of the original read by an image reader or the like, in order to obtain the binarization result in the edge enhancement processing which determines the presence / absence and the direction of the edge enhancement based on the result. Considering image data processing using a predetermined matrix, the following is understood. That is, considering the detection of contours of characters and line images, it is preferable to use a large matrix such as a 5 × 5 matrix. This is because even if a small matrix such as a 3 × 3 matrix is used, local density variations are detected. On the other hand, in the process of detecting the presence / absence and the direction of edge enhancement from the pattern of the binarization result, the size of the matrix is preferably as small as a 3 × 3 matrix. This is because a pattern matching ROM table for detection becomes huge when a large matrix such as a 5 × 5 matrix is used.

In the edge emphasis processing of the present invention, based on the above consideration, in order to execute more accurate edge emphasis processing,
Two types of matrices, large and small, of 5 × 5 pixels and 3 × 3 pixels each having a target pixel are used. Basically,
Using a 5 × 5 matrix, a threshold value is obtained from the average value of the image data of the pixels forming the matrix, and binarization processing is executed. Next, using a 3 × 3 matrix, the presence / absence of execution of the edge enhancement processing and the detection of the edge enhancement direction are performed based on the binarization result. When the edge emphasis processing is executed, the edge emphasis processing is executed for the pixel of interest using a Laplacian filter formed of a 3 × 3 matrix corresponding to the detected edge emphasis direction.

However, if the binarized result by the 5 × 5 matrix satisfies a predetermined condition (in the case of this embodiment, all are black or white), more detailed binarized data is obtained. Then, a threshold value is obtained from the average value of the image data of the pixels forming the 3 × 3 matrix, and binarization processing is executed. Subsequently, using a 3 × 3 matrix, the presence / absence of execution of the edge enhancement process and the detection of the edge enhancement direction are performed based on the binarization result. Here, when the edge emphasis processing is executed, the edge emphasis processing is executed for the pixel of interest using a Laplacian filter composed of a 3 × 3 matrix corresponding to the detected edge emphasis direction.

The edge enhancement processing of the present invention obtains a simple one-dimensional Laplacian output value for a plurality of directions, and compares it with the conventional edge enhancement processing in which one of them is selected. Excellent in terms. First, since the binarization processing and the pattern matching are performed at the time of detecting the direction in which the edge is emphasized, there are few erroneous detections caused by the comparison based on the simple magnitude relationship as in the conventional art, and the circuit configuration is simple. become. Secondly, depending on the binarized pattern, edge enhancement is not performed, so that more flexible handling is possible.

(2) Configuration of Image Processing Device FIG. 2 shows the overall configuration of the image processing device of the present invention. The image processing apparatus receives the image data read by the image reader, forms a 5 × 5 matrix, and stores the memory 1, the edge direction selection unit 20, the four-direction Laplacian calculation unit 9, and the four-direction matrix. Select Laplacian from Laplacian in the direction of executing edge enhancement processing. 4 directions → 1
The direction Laplacian selection unit 10 and the edge emphasis amount calculation unit 11 that adds the Laplacian calculation result output from the selection unit 10 to the image data read by the image reader as an edge emphasis amount and outputs the result.

The image data input to the memory 1 is CC
Consider the multi-bit serial image data after the output value from D is digitally converted. In addition, the edge direction selection unit 2
At 0, by pattern matching after the binarization processing,
The presence or absence of edge enhancement and the direction are detected. On the other hand, the four-direction Laplacian calculator 9 calculates a total of four Laplacian directions including horizontal, vertical, and diagonal two directions. In the 4-direction to 1-direction Laplacian selection unit 10, the Laplacian in the edge direction selected by the edge direction selection unit 20 is selected from the output of the 4-direction Laplacian calculation unit 9 and is used as the edge emphasis amount in the edge emphasis amount calculation unit 11. Output. The edge emphasis amount calculation unit 11 adds the input edge amount to the value of the image data of the target pixel. Through the above processing, the image data subjected to the edge enhancement processing is output from P _OUT .

Here, the edge direction selection unit 20 is set to 5
Based on the threshold value output from the 5 × 5 threshold value calculation unit 2 and the threshold value output from the calculation unit 2, the average density value of each pixel in the × 5 matrix is obtained and is output as the threshold value. A first binarization unit 4 which binarizes each pixel in the 5 × 5 matrix, a binarization result, and a preset binarization pattern are compared to detect a corresponding pattern. 1 pattern matching unit 6 and 3 × 3 threshold value calculation unit 3 that obtains an average density value of each pixel in a 3 × 3 matrix and outputs this as a threshold value, and outputs from the calculation unit 3 2 pixels for each pixel in the 3 × 3 matrix based on the threshold
Second binarization unit 5 for binarization, second binarization result, and second binarization result set in advance to detect a corresponding pattern, and second pattern matching unit 7 and 5 × 5 matrix The data selector 8 which compares and examines the calculation result of 3 and the calculation result of the 3 × 3 matrix, and outputs either one of the results.
Composed of and.

(3) Detailed Description of Each Functional Unit FIG. 3 is a configuration circuit diagram of the 5 × 5 matrix memory 1. The memory 1 is composed of four 8-bit FIFOs and 17 8-bit D-FFs. As shown in FIG. 4, this corresponds to five pixels (Pi, j: i = 1 to 5, j in the main and sub-scanning direction with the target pixel P33 to which the edge enhancement processing is applied as the center.
= 1 to 5) is a memory for referring to the image data at one time. Incidentally. In the circuit of the memory 1, the image data is carried in synchronization with the CLK signal (clock signal). Therefore, all the Pin and P11 to P55 image data are 8-bit data.

Next, before explaining each part of the edge direction selection unit 20, the direction in which the edge is emphasized is finally determined from the binarization result by the two types (5 × 5 and 3 × 3) of the matrix. How to do is described.

FIG. 5A is a diagram showing a state of image data in a predetermined area when a typical pencil-written original is read. Here, the density of each pixel in the drawing increases in the order of one pixel with diagonal lines, two pixels with diagonal lines, and cross-hatched pixels. In FIG. 5B, the pixel marked with “*” is the pixel of interest, and 5 × 5 pixels are formed around this pixel.
The pixels in the matrix are binarized. Also,
In FIG. 5C, the pixel marked with “*” is the pixel of interest, and the pixels in the 3 × 3 matrix formed around this pixel are binarized. The threshold value for binarization used in this case uses the average value of the density values of each pixel in each matrix.

As can be understood by looking at the binarization result, 5
Binarization result in x5 matrix (see FIG. 5 (b))
From this, it is clear that there are edges in the vertical direction, but the binarization result in the 3 × 3 matrix (see FIG. 5).
It cannot be said from (c)) that this is always the case. In this way, for a low-frequency image in which one cycle of light and shade exists for 10 pixels, a 2 × 5 × 5 matrix is used.
It is better to digitize. This is because the 3x3 matrix is too sensitive to local concentration changes.

On the other hand, the read image (a) shown in FIG. 6 is an image in which light and shade are present in one cycle for 5 pixels. Here, the density of each pixel in the figure is one shaded line, two shaded lines, and “X”.
The pattern becomes darker in the order of cross hatching and diamond cross hatching. For such an image, FIG.
When the Laplacian filter shown in FIG. 6 is used, the range to which the edge emphasis is applied is as shown in FIG. However, when the read image (a) is binarized, the read image has moire, and the result is as shown in FIG.
Considering that the binarization result of FIG. 6B is almost the same as the binarization result of each pixel with a 5 × 5 matrix, the presence / absence of edge enhancement from the 3 × 3 matrix pattern centered on the target pixel is considered. Is obtained, the area to which the edge enhancement is applied becomes the shaded area in FIG. 7B. As described above, the edge enhancement using the binarization result of the 5 × 5 matrix may work in the direction of enhancing the moire in the image having the moire.

As for the area without hatching in FIG. 7B,
The direction of edge enhancement can be obtained by using the binarization result of the 3 × 3 matrix. Thus, in the present invention, 3 ×
The binarization result of 3 matrices and 5 × 5 matrix is used as follows. First, the presence / absence of edge enhancement and its direction are obtained from the binarization result using a 5 × 5 matrix, and if there is edge enhancement, that direction is finally selected. Next, if the binarization result using the 5 × 5 matrix is all white or all black, the presence / absence of edge enhancement and its direction are obtained by referring to the binarization result using the 3 × 3 matrix. Here, with edge emphasis,
If that direction is finally selected and there is no edge enhancement, no edge enhancement is performed. From the binarization result using the 5 × 5 matrix, when it is determined that there is no edge enhancement and neither all white nor all black, it is determined that the target pixel does not have edge enhancement.

By the way, of the binarization results using the 5 × 5 matrix, a sufficient discrimination result can be obtained by referring to only the pattern of the central 3 × 3 matrix. Therefore, in the image processing apparatus of the present invention, the size of the matrix for obtaining the threshold value for binarization is made larger than the size of the matrix used for the reference of the subsequent pattern, thereby reducing the scale of the hardware circuit. .

FIG. 8 is an internal block diagram of the 5 × 5 matrix threshold calculation unit 2 and the 3 × 3 matrix threshold calculation unit 3. The calculation unit 3 is composed of a total of 19 average value circuits. Each average value circuit has two input terminals and one output terminal, and the density value of the two pixels input from each input terminal is calculated. The average value is output from the output terminal. Only the upper 6 bits of each image data are used for this calculation.

In order to calculate the threshold value for binarization in the 3 × 3 matrix threshold value calculation unit 3, the density values of the pixels indicated by diagonal lines in FIG. 9A are input to the input terminals. Then
The average value AVR33 is calculated.

Further, the 5 × 5 matrix threshold value calculation unit 2
In order to calculate the threshold value for binarization with, the density values of the pixels shown by diagonal lines in FIG. 9B are input to the input terminals and the average value AVR55 is calculated.

Next, FIG. 10 shows the first binarization unit 4 and the first binarization unit 4.
3 is a circuit configuration diagram of a pattern matching unit 6. FIG. The circuit includes nine comparators, a ROM that performs pattern matching processing based on the comparison result output from each comparator,
It is composed of a logic circuit including two inverters and four OR gates. 5 × for each input terminal of the comparator
The density value of each pixel in the 5 matrix is input, and the threshold value AVR55 calculated by the 5 × 5 threshold value calculator 2 described above is input to all of the density values. The threshold value is compared with the density value of each pixel, binarization processing is performed based on the comparison result, and the result is stored in the ROM.
Is output to the address. The ROM performs pattern matching processing based on the input binarization result and outputs this result. The information signal output from the ROM is a signal indicating the presence or absence of edge enhancement (EDGEON55) and a signal indicating the direction of edge enhancement (horizontal: EHON55, vertical: EVON5).
5, diagonal-1: ED1ON55, diagonal-2: ED2ON55) and a switching signal (E that indicates that the binarization results are all white or all black
DGESEL).

FIG. 11 is a circuit configuration diagram of the second binarization unit 5 and the second pattern matching unit 7. The circuit includes nine comparators, a ROM that performs pattern matching processing based on the comparison result output from each comparator,
It is composed of a logic circuit including two inverters and four OR gates. 3x for each input terminal of the comparator
The density value of each pixel in the three matrices is input, and the threshold value AVR33 calculated by the 3 × 3 threshold value calculation unit 4 described above is input to all the density values. The input threshold value is compared with the density value of each pixel, binarization processing is performed based on the comparison result, and the result is output to the address of the ROM. The ROM performs pattern matching processing based on the input binarization result and outputs this result. The information signal output from the ROM is a signal indicating the presence or absence of edge enhancement (EDGEON33) and a signal indicating the direction of edge enhancement (horizontal: EHON3
3, Vertical: EVON33, diagonal-1: ED1ON33, diagonal-2: ED2ON3
3) and are respectively output from the output terminals.

The following FIGS. 12, 13 and 14 show the ROM
FIG. 3 is a diagram showing patterns of a 3 × 3 matrix stored in a table classified according to the direction in which edges are emphasized (oblique-1 direction, oblique-2 direction, vertical direction, horizontal direction). The binarization result of the 3 × 3 matrix input to the ROM is shown in FIG.
It is collated with each pattern shown in FIGS. 13 and 14.
Here, for example, when the input binarization result data corresponds to one of the edge enhancement patterns in the vertical direction, RO
From M, EVON having a value of 1 is output, and EHON, ED1ON, and ED2ON having a value of 0 are output. Further, as a result of collating the binarized results of the input 3 × 3 matrix, if there is no corresponding pattern, the ROM table is set so that the edge emphasis is not executed.

In the data selector 8, according to the EDGESEL signal obtained by the first pattern matching section 6, the first pattern
When the binarization result in the 5 × 5 matrix in the pattern matching unit 6 is all white or all black, the binarization result when using the 3 × 3 matrix (EDGEON33, EHON
33, ED1ON33, ED2ON33), and if they are not all white or all black, the binarization result (EDGEON55, EHON55, EVON55, ED1ON55, ED2ON) when using a 5x5 matrix
55) is selected and output from the output terminal as EDGEON, EHON, ED1ON, ED2ON.

FIG. 15 is an internal block diagram of the 4-way Laplacian calculator 9. Here, the output values (EDGE-H, EDGE-V, EDGE-D1, EDGE-D2) of the four-dimensional one-dimensional second-order differential filter of horizontal, vertical, and diagonal two directions shown in FIG. 16 are calculated. In the 4-direction to 1-direction Laplacian selection unit 10, 4 according to the signals EDGEON, EHON, EVON, ED1ON, ED2ON indicating the presence or absence of the edge enhancement and the direction obtained by the data selector 8,
Output of directional Laplacian calculator 9 EDGE-H, EDGE-V, EDGE-D
Select one from EDGE-D2 and output. When the edge enhancement is not performed, the output of the 4-direction → 1-direction Laplacian selection unit 10 is reset to zero.

The edge enhancement amount (E
DGE) is input to the edge emphasis amount calculation unit 11, and the pixel of interest
8-bit image data Pout that has been added to P33 and subjected to edge enhancement processing is obtained. FIG. 17 shows a block diagram of the configuration of the edge emphasis amount calculation unit 11. In this circuit, the addition result takes the upper and lower limits (00 and ff) of the image data.
So, I am doing the limit.

By performing the edge enhancement processing as described above, it is possible to eliminate the edge enhancement in unnecessary directions with a simple structure. As a result, it is possible to eliminate faint characters and partial thickening due to repeated copying operations.
In particular, when the copying is repeated a plurality of times, the storability of the character image is improved.

Further, in this embodiment, the edge direction selection unit 2
At 0, the presence / absence of execution of the edge emphasis processing and its direction are checked, but the degree of execution of the edge emphasis processing, that is, the control of the edge amount (0 when the edge emphasis processing is not executed)
Alternatively, the direction may be checked.

In this embodiment, the binarization pattern by the small (3 × 3) matrix is referred to only when the binarization results by the large (5 × 5) matrix are all white or black. However, this binarized pattern may be slightly different from the pattern that is all black or white.

[0048]

Since the image processing apparatus of the present invention determines the presence / absence of the edge enhancement and the direction thereof based on the binarized pattern, the edge enhancement in the unnecessary direction can be performed with a circuit configuration simpler than the conventional one. Can be eliminated. As a result, it is possible to eliminate the blurring of the characters and the partial thickening due to the copy operation, and it is possible to improve the storability of the character image particularly when copying is repeated a plurality of times.

[Brief description of drawings]

FIG. 1 is a diagram showing a binarization result when a simple binarization process is performed with a predetermined threshold value and an example of a case where edge enhancement is performed based on the result.

FIG. 2 is a schematic configuration diagram of an image processing apparatus according to the present invention.

FIG. 3 is a configuration circuit diagram of a 5 × 5 matrix memory 1.

FIG. 4 is a 5 × 5 formed by the memory 1 shown in FIG.
It is a figure which shows the matrix of.

FIG. 5A is a diagram showing a state of a predetermined area in image data obtained by reading a typical original document written in pencil. (B) is a diagram showing a result of binarization using a 5 × 5 matrix with a “*” mark as a target pixel.
(C) is a diagram showing a result of binarization with a 3 × 3 matrix with a “*” mark as a target pixel.

FIG. 6A is a diagram of image data in which light and shade of one cycle exists for 5 pixels. (B) is a diagram showing a range in which the edge emphasis processing is mainly performed in the image data of (a) when the Laplacian filter is used.

7A is a diagram not showing a result of binarizing the read image of FIG. 6A, and FIG. 7B is a range relating to edge enhancement processing according to the binarization result of FIG. I asked for.

FIG. 8: 5 × 5 matrix threshold value calculators 2 and 3 ×
3 is an internal configuration circuit diagram of a 3-matrix threshold value calculation unit 3. FIG.

9A is a diagram showing pixels of a 3 × 3 matrix input to the circuit of FIG. 8 by using diagonal lines, and FIG. 9B is a diagram showing FIG.
It is a figure which shows the pixel of a 5x5 matrix input into the circuit of FIG. 8 using a diagonal line.

FIG. 10 is a circuit configuration diagram of a binarization unit and a pattern matching unit 5.

11 is a circuit configuration diagram of the binarization unit 6 and the pattern matching unit 7. FIG.

FIG. 12 is a diagram showing a 3 × 3 matrix pattern stored in the ROMs of the pattern matching units 5 and 7 for performing edge enhancement in an oblique direction.

FIG. 13 is a diagram showing a 3 × 3 matrix pattern for edge enhancement in the vertical direction, which is stored in the ROMs of the pattern matching units 5 and 7.

FIG. 14 is a diagram showing a 3 × 3 matrix pattern which is stored in the ROMs of the pattern matching units 5 and 7 and which performs edge enhancement in the horizontal direction.

FIG. 15 is an internal block diagram of a four-way Laplacian calculator 9.

16 is a four-way Laplacian calculator 9 shown in FIG.
It is a figure which shows what calculated | required the output value of the vertical, horizontal, and diagonal one-dimensional one-dimensional secondary differential filter output from.

FIG. 17 is a diagram showing a configuration block of an edge enhancement charge adding unit 11.

18A and 18B are diagrams showing a typical Laplacian filter.

19A to 19C are diagrams showing an example in which edge enhancement processing is performed using the Laplacian filter shown in FIG. 18A.

[Explanation of symbols]

 1 ... 5 × 5 matrix memory 2 ... 5 × 5 threshold calculation section 3 ... 3 × 3 threshold calculation section 4,5 ... binarization section 6,7 ... pattern matching section 8 ... data selector 9 ... 4 directions Laplacian calculation unit 10 ... 4 directions → one direction Laplacian selection unit 11 ... Edge enhancement amount calculation unit

─────────────────────────────────────────────────── ───

[Procedure amendment]

[Submission date] January 10, 1994

[Procedure Amendment 1]

[Document name to be amended] Statement

[Name of item to be corrected] Fig. 18

[Correction method] Change

[Correction content]

18A to 18C are diagrams showing a typical Laplacian filter.

Claims (5)

[Claims] 1. An image processing apparatus for executing edge enhancement processing of image data of a read original image, which is binary based on a value of image data of pixels forming a first matrix having a pixel of interest at the center. First deciding means for deciding a threshold for performing binarization, and first binarization for binarizing the image data in the first matrix using the threshold decided by the first deciding means. Means, a judgment means for judging the presence or absence of edge enhancement processing based on the binarized data by the first binarizing means, an edge amount detecting means for detecting the edge amount in the first matrix, and the judging means. Image processing, characterized in that, when it is determined that the edge emphasis processing is to be executed, the edge emphasis means executes the edge emphasis processing for the pixel of interest based on the detection result by the edge amount detection means. Location. 2. An image processing apparatus for executing edge enhancement processing of image data of a read original image, which is binary based on the value of image data of pixels forming a first matrix having a pixel of interest at the center. First deciding means for deciding a threshold for performing binarization, and first binarization for binarizing the image data in the first matrix using the threshold decided by the first deciding means. Means, a first edge amount detecting means for detecting an edge amount in the first matrix, and a pixel of interest in the center, and image data of pixels forming a second matrix smaller in size than the first matrix. Second determining a threshold for binarization based on the value
Deciding means, second binarizing means for binarizing the image data of the pixels forming the second matrix by using the threshold value by the second deciding means, and detecting the edge amount in the second matrix Second edge amount detecting means, judging means for judging whether or not the binarized data by the first binarizing means satisfies a predetermined condition, and the binarizing data by the judging means satisfies the predetermined condition. When it is determined that the edge enhancement processing is not performed, the first determination means determines whether or not the edge enhancement processing is performed based on the binarized data by the first binarization means, and the first determination means determines to execute the edge enhancement processing. , Based on the detection result by the first edge amount detecting means,
A first edge enhancing means for performing edge enhancing processing on a pixel of interest of the first matrix; and, if the determining means determines that the binarized data satisfies a predetermined condition, the second binarizing means performs binary processing. Second judgment means for judging the presence or absence of the edge emphasis processing based on the digitized data, and when the second judgment means judges to execute the edge emphasis processing, based on the detection result by the second edge amount detection means,
An image processing apparatus comprising: a second edge enhancement unit that performs edge enhancement processing on a pixel of interest of a second matrix. 3. An image processing apparatus for executing edge enhancement processing of image data of a document image read by an image reader, wherein the image processing apparatus is based on image data values of pixels forming a first matrix having a pixel of interest at the center. Determining means for determining a threshold value for binarizing the image data, and binarizing means for binarizing the image data in the first matrix using the threshold value determined by the determining means; Determination means for determining the presence or absence of edge enhancement processing based on the binarized data in the second matrix having a smaller size than the matrix, and edge amount detection means for detecting the edge amount in the second matrix, An edge enhancement unit that executes the edge enhancement process based on the detection result of the edge amount detection unit when the edge enhancement process is determined to be performed by the unit. That image processing apparatus. 4. The claim 1, claim 2, or claim 3.
In the image processing device described in, the judgment means includes a memory for storing a plurality of binarized patterns, and compares the binarized data by the binarized means with the pattern stored in the memory. An image processing apparatus, which determines whether or not to perform edge enhancement processing based on the above. 5. The image processing apparatus according to claim 1, claim 2, claim 3, or claim 4, wherein the determination means further determines the direction of the edge enhancement processing, and the edge amount detection means includes a plurality of Detecting the edge amount in the direction, the edge emphasizing unit selects the edge amount in the direction judged by the judging unit from the edge amounts in the plural directions detected by the edge amount detecting unit, and executes the edge emphasizing process. An image processing device characterized by: