JP3804590B2 - Image processing device - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to addition of a specific mark in original image data.
[0002]
[Prior art]
One direction of image quality improvement in an image forming apparatus such as a copying machine is to improve image reproducibility by performing image processing according to the type of document. Here, it has been proposed to record a specific mark or the like on a sheet, and to control image processing based on information obtained by reading the mark.
[0003]
For example, in Japanese Patent Laid-Open No. 2001-189865, a copy prohibition mark or bar code is recorded on a sheet in order to prohibit illegal copying. When the image processing circuit detects a copy prohibition mark or barcode in the image data, appropriate processing is performed based on the obtained information (copy prohibition, etc.). In JP-A-8-305791, a mark is formed on a sheet, and information is obtained from the mark. The information provided by the mark is various control information such as a document direction and a scanning direction. Here, by forming marks at the four corners of the paper, the same recognition processing method can be used regardless of how the original is placed.
[0004]
[Problems to be solved by the invention]
As described above, it has been proposed to form an image by adding a specific mark representing information on the image to the original image. For example, in a so-called generation copy in which a copied material is copied as a document, a technique in which an operator manually switches image processing is used in order to improve image reproducibility. In the case of the generation copy, for example, the character tends to become thin or sluggish depending on the MTF characteristics. Therefore, the reproducibility of the character image can be improved by performing an expansion process or performing an image process that emphasizes the edge more than other cases. In view of this, it is conceivable for generation copies to add a specific mark that is invisible to human eyes to the original image so that it can be automatically determined whether or not the copied one is a manuscript. In this way, it is possible to automatically perform image processing suitable for generation copying only by reading a document image. Here, it is preferable that the specific mark itself does not affect the original image, and it is desired that the specific mark be small and difficult to be identified by the human eye. On the other hand, it is necessary to accurately detect the specific mark. Further, it is desirable to be able to effectively add the specific mark to the image at the time of image formation and identify the specific mark at the time of image reading.
[0005]
An object of the present invention is to effectively identify a specific mark when reading an image.
[0006]
[Means for Solving the Problems]
An image processing apparatus according to the present invention includes: a discriminating unit that discriminates a specific mark having a predetermined color and shape; a first count unit that counts the number of specific marks discriminated by the discriminating unit; Second counting means for counting the number, and when the count result of the first counting means is not less than the first threshold value and the count result of the second counting means is not more than the second threshold value, Determination means for validating the determination by the determination means.
Preferably, further comprising third counting means for counting the number of pixels of the background color of the specific mark, wherein the determination means is configured such that when the count result of the third counting means is equal to or greater than a third threshold value, The discrimination by the discrimination means is valid. Thus, in an apparatus for recognizing a specific mark formed with a specific color and shape, a specific pixel color (mark color or background color) within a detection range is used as a determination condition as to whether or not a specific mark exists in the input image. By adding an existing ratio, it is possible to prevent a general image from being erroneously recognized as a mark. For example, if it is premised that a mark exists on a specific background color, the mark is not determined as a mark because the image is considered a general image when the ratio of the background color is small. Similarly, when the ratio of the specific color is large, the mark is not determined.
[0007]
In the image processing apparatus, preferably, the count result of the first count unit is equal to or less than a fourth threshold value different from the first threshold value, or the count result of the second count unit is the second threshold value. If the threshold value is not more than the fifth threshold value or the count result of the third counting means is not more than the sixth threshold value that is different from the third threshold value, the determination by the determining means is invalidated. .
[0008]
A second image processing apparatus according to the present invention includes: a discriminating unit that discriminates a specific mark having a predetermined color and shape; a first counting unit that counts the number of specific marks discriminated by the discriminating unit; Third counting means for counting the number of pixels of the background color, and determination means for determining that the determination by the determination means is valid when the count result of the third counting means is equal to or greater than a third threshold value. .
[0009]
Preferably, in the image processing apparatus, the determination of the specific mark by the determination unit and the count of the number of pixels of the color of the specific mark by the second count unit are the same as the specific mark from the input image by the color extraction unit. This is performed in parallel following the extraction of color pixels.
[0010]
Preferably, in the above-described image processing apparatus, in order to recognize a specific mark more accurately, a range of pixels that are targets of pattern matching of the specific mark is divided into a plurality of central blocks including the pixel of interest and a plurality of blocks around it. In each block, the number of pixels of a specific mark color is counted in each block. If the count value in each block satisfies a predetermined condition, it is determined that the mark is a specific mark.
[0011]
DETAILED DESCRIPTION OF THE INVENTION
Hereinafter, embodiments of the present invention will be described with reference to the accompanying drawings.
The generation copy (generation copy) is to perform copying using a copied copy as a manuscript. In the present invention, when a document is printed by a color copying machine or a multi-function multifunction device, a pattern made up of marks of a specific shape is used so that it can be determined whether the copied document is a document. A copy is made by adding to the original image. When an image is read using a copy to which a pattern composed of a plurality of specific marks is added as an original and an image is formed on a sheet, the pattern of the specific mark can be detected from image data obtained by reading the original. Thereby, it can be automatically determined that the copy is a generation copy. When it is determined that the copy is a generation copy, a copy is created by performing image processing suitable for such a document. For example, deterioration of the character image can be prevented and reproducibility can be improved by emphasizing the edge more than usual, eliminating character shakiness, or correcting the color. Such a specific mark pattern can also be used to provide information other than the generation copy. However, only the application to the generation copy will be described below, so the specific mark is abbreviated as a generation copy mark (GC mark). It is also said. The specific mark pattern is, for example, an equidistant arrangement of a predetermined number of specific marks.
[0012]
The generation copy process will be described with reference to FIGS. The original to be copied for the first time naturally does not include a pattern consisting of generation copy marks (hereinafter referred to as GC marks). When such a document is read (S10) and the obtained image data is subjected to GC mark detection processing (S12), it is determined that there is no GC mark pattern (NO in S14). Therefore, normal image processing is performed on the image data (S16). Next, a process of adding a GC mark pattern to the margin of the paper is performed (S20). Then, an image is formed on the paper (S22).
[0013]
In the copy thus created (first generation copy), a pattern composed of a plurality of GC marks is formed in the margin of the paper. FIG. 2 schematically shows an example. Margins are provided on four sides of the paper, and a plurality of GC mark patterns are formed along the sides in the margins. The GC mark has a flat shape and is formed to be horizontally long in the direction along the side of the sheet. As a pattern composed of a plurality of GC marks, in this example, a plurality of GC marks are arranged at equal intervals. The color of the specific mark is an inconspicuous color, for example, yellow.
[0014]
Next, when copying the first generation copy thus obtained as a document, the document is read (S30), and a GC mark detection process is performed on the obtained image data (S32). A mark pattern is detected (YES in S34), and it is determined that the copy is a generation copy. In that case, image processing for generation copy is performed on the image data (S38). Next, a process of adding a GC mark pattern to the margin is performed on the image data (S40). Then, an image is formed on the paper (S42).
[0015]
Here, the following points should be considered for the specific mark. When a specific mark is recognized from the input image, the mark itself needs a certain size for pattern matching. When a specific mark is printed in the margin portion, a large margin space is required to print a large mark in consideration of ease of recognition and influence on the image. Further, when an input image is stored for recognition in pattern matching, the capacity of the line memory for storing the input image also increases because a storage capacity larger than the specific mark is required. In addition, if the specific mark is large, the mark is easily noticeable even if an inconspicuous color is adopted. In addition, when a specific mark is printed on only one side, there is a problem that the mark position changes depending on how the original is placed on the platen glass of the scanner, or it takes time to recognize from the scanning operation. .
[0016]
Therefore, in the present invention, the following specific marks are used as shown in FIG.
(1) The shape of the specific mark is a horizontally long shape (for example, a rectangle), and is recorded horizontally along the side of the paper. Since the specific mark is identified by pattern matching, it needs to have a certain size, but if it is too large, it will stand out. Since a horizontally long mark is used, printing can be performed in a small margin area while ensuring a certain mark size, so that a small margin area is required. For this reason, the influence on the document image can be reduced. In addition, since a horizontally long mark is used, it can be made inconspicuous even if the mark area is the same. Further, since the horizontally long mark is used, the storage capacity of the line memory in the processing circuit can be reduced, and the configuration of the processing circuit can be simplified.
[0017]
(2) A plurality of specific marks are printed along the side of the paper in the margin of the paper. Since a plurality of marks are recorded, misrecognition can be avoided even if there are some defects. Since a plurality of marks are arranged along the side, printing can be performed in a small margin area. Further, as shown in FIG. 4, since the number of lines required for pattern matching is small, the capacity of the line memory in the processing circuit can be reduced. In one example, the interval between the specific marks is constant.
[0018]
(3) The specific mark is printed on each of the four sides of the paper. The detection range of the specific mark is, for example, a range of the first designated number of lines at the time of scanning, and if necessary, a designated range in the main scanning direction. Therefore, as shown in FIG. 5, the mark can be recognized by performing the same processing in the same sub-scanning direction range regardless of how the document is placed. Therefore, the configuration of the processing circuit can be simplified, and the processing time until recognition can be shortened.
[0019]
FIG. 6 shows the overall configuration of a color copying machine that appropriately processes the above-mentioned generation copy. The configuration of this copying machine is the same as that of the prior art except for the configuration of the specific document recognition circuit 24 and the GC mark related processing.
[0020]
In the scanner, the reflected light from the original read by the CCD sensor 10 is converted into digital RGB read data by the AD converter 12. The read data is subjected to various correction processes such as shading correction, line-to-line correction, and chromatic aberration correction in the read correction unit 14. The corrected read data is processed in two paths.
[0021]
On the other hand, preprocessing for image processing is performed by the resolution conversion (enlargement / reduction) unit 16, the RGB gamma correction unit 18, the brightness / color difference separation unit 20, and the GC mark removal unit 22. The resolution conversion (enlargement / reduction) unit 16 performs a scaling process if necessary. The RGB gamma correction unit 18 performs gamma correction on the RGB data. The HVC converter 20 converts the RGB data into lightness V and color differences Cr and Cb. The GC mark removing unit 22 removes the GC mark from the blank portion when the GC mark is detected from the read data.
[0022]
On the other hand, whether or not the corrected read data is a specific document such as a banknote is determined in the specific document recognition circuit 24. Further, a specific mark recognition chip 26 for identifying the GC mark can be attached to and detached from the specific document recognition circuit 24. The specific mark recognition chip 26 includes a CPU 100 that controls the whole, a ROM 102 that stores a recognition program, a RAM 104 that is a work area, and a processing circuit 106 that performs identification processing under the control of the CPU 100.
[0023]
Next, image adjustment is performed according to the characteristics of the image. The sharpness correction unit 28 adjusts the sharpness of the image, and the HVC adjustment unit 30 independently adjusts brightness, saturation, and hue. Then, the color space conversion unit 32 performs conversion of the adjusted data into density data, undercolor removal / inking process, conversion to CMYK data, and the like. Although not shown, color / monochrome original automatic discrimination and automatic background level adjustment are also performed in parallel. If the specific document or specific mark is not detected by the specific document recognition circuit 24 or the specific mark recognition chip 26, normal image adjustment is performed. If a specific mark is detected, image adjustment for generation copy is performed. Is done. Thereafter, a specific mark (GC mark) is added to the margin by the addition processing unit 34. Although not shown, a binarization process is performed for a monochrome image.
[0024]
Next, the JPEG encoding processing unit 36 compresses the data. The compressed data is sent to the printer unit 40 or the external device 42 via the scanner image IF 38. The printer unit 40 forms an image on a sheet based on the received data. Here, a pattern composed of a plurality of specific marks is formed on the four sides of the sheet.
[0025]
The specific mark recognition chip 26 recognizes a pattern composed of a plurality of specific marks arranged at equal intervals in the main scanning direction. In recognition of such a specific mark pattern, it is necessary to detect specific marks arranged at predetermined intervals from image data obtained by reading a document. When recognizing a specific mark composed of a plurality of mark elements, it is difficult to accurately detect the number and arrangement of mark elements if colors and shapes similar to the mark elements are mixed in a general image.
[0026]
Therefore, in the present embodiment, a general image is not misidentified as a mark for a mark on a base of a specific color. Specifically, a general image is marked by adding a ratio of the presence of a specific pixel color (mark color or background color) within the detection range to the determination condition as to whether or not a specific mark exists in the input image. Prevent misrecognition. For example, if it is premised that a mark exists on a specific background color, if the background color ratio is small, it is considered as a general image and is not determined as a mark. Similarly, the mark is not determined when the specific color ratio is large.
[0027]
In a copying machine that performs copying by adding an identification mark to a document image, the ratio of the number of pixels in the mark color is inspected to determine whether or not a mark exists in Japanese Patent Application Laid-Open No. 4-222644. Proposed. However, in this case, if there is a pixel with the same color but not a mark, an erroneous determination is made. In the present embodiment, as shown in FIG. 7, the mark detection itself is performed by another pattern matching circuit, and the inspection for the background color pixel and the pixel for the mark color are also performed. And it is judged using those results whether the mark is erroneously detected.
[0028]
Specifically, in addition to the first counting means for counting the number of specific marks, the second counting means for counting the number of pixels of the color of the specific mark, and the first counting means for counting the number of pixels of the background color of the specific mark. 3 counting means. When the count result of the first count unit is equal to or greater than the first threshold value and the count result (mark color ratio) of the second count unit is equal to or less than the second threshold value, The determination that it exists is valid, but otherwise it is invalid. Alternatively, when the count result of the first count means is equal to or greater than the first threshold value and the count result (background color ratio) of the third count means is equal to or greater than the third threshold value, The determination that it exists is valid, but otherwise it is invalid. More preferably, the count result of the first count means is equal to or greater than the first threshold value, and the count result (mark color ratio) of the second count means is equal to or less than the second threshold value. When the count result (background color ratio) of the third count means is equal to or greater than the third threshold value, the determination that the mark is a specific mark is valid, but otherwise is invalid.
[0029]
Furthermore, any of the following conditions may be added. That is, the count result (number of marks) of the first count means is equal to or less than a fourth threshold value different from the first threshold value, or the count result (mark color ratio) of the second count means is A sixth threshold value that is equal to or less than a fifth threshold value that is different from the second threshold value, or whose count result (background color ratio) of the third counting means is different from the second threshold value. If it is below, the determination that the mark is a specific mark is invalidated. That is, if any of these conditions is satisfied, the discrimination from the specific mark is invalidated even if the other conditions are satisfied.
[0030]
The mark filter for detecting the specific mark has a size of 11 × 11 pixels including the target pixel. In pattern matching of a specific mark, for more accurate mark recognition, in consideration of misrecognition due to the shape of the specific mark, an area of 11 × 11 pixels is divided into a plurality of central blocks including the pixel of interest and a plurality of peripheral blocks. And the number of pixels of a specific mark color is counted in each block. If the count value in each block satisfies a predetermined condition, it is determined that the mark is a specific mark.
[0031]
Hereinafter, recognition of a pattern composed of a plurality of specific marks by the specific mark recognition chip 26 will be described. FIG. 8 shows a flow of specific mark recognition by the CPU 100 in the specific mark recognition chip 26. First, R, G, B image data (8 bits each) is input (S100). If necessary, perform pre-processing such as scaling. Next, the color of the GC mark is extracted from the input image data (S102). Here, binarization is performed by determining whether or not the input image density of R, G, and B is within a predefined reference density range. Thereby, binary data is obtained. Next, noise is removed from the binary image data (S104). Here, an isolated point consisting of one pixel is removed. For the binary data obtained in this way, a specific mark element shape is detected using mark detection parameters for pattern matching (S106). Here, for example, the mark element shape is extracted using an element filter of 11 × 11 pixels. Next, it is inspected whether or not the detected mark elements are at a predetermined interval (S108). Here, if the difference between the previously detected mark element and the x coordinate position is a natural number multiple of the specified mark interval, the mark element is determined to be a mark element. Next, the number of mark elements at a predetermined interval is calculated (S110). Next, the number of mark elements is summed to determine whether or not the number is a specified number, and the result is output (S112). Here, if the number of mark elements is a specified number, it is determined that the copy is a generation copy. The GC mark recognition process in the specific mark recognition chip 26 has been described as a software process by the CPU 100. However, this is for convenience of explanation, and in practice, at least a part of the process is configured as a dedicated hardware circuit. it can.
[0032]
Further, the mark color pixel is processed in parallel for the binary image obtained in step S102. That is, the mark color pixels within the mark detection range are counted (S120). Next, the mark color pixel area ratio (nMark / nAll) is calculated and compared with a preset threshold value MarkRatioMin. If the mark color pixel area ratio is larger than the threshold value, the image is too large than the specified mark. It is determined that it is a (general image) (S122). When it is determined that the image is a general image, a detection result signal indicating that the image is not a mark is output (S124).
[0033]
Further, a background (for example, white) is extracted from the input image data (S130). Here, binarization is performed by determining whether or not the input image density of R, G, and B is within a predefined reference density range. As a result, binary data for the background is obtained. Next, the background color pixels within the mark detection range are counted (S132). Next, the background color pixel area ratio (nWhite / nAll) is calculated and compared with a preset threshold WhiteRatioMin. Image) (S134). When it is determined that the image is a general image, a detection result signal indicating that the background color is not used is output (S136).
[0034]
Then, a detection result signal indicating whether or not the number of mark elements output in step S112 is a specified number, a detection result signal indicating that the mark element is not a mark from step S124, and a detection result signal indicating that the number is not a background color from step S136 are provided. In response, it is determined whether or not a mark exists. Here, even if a detection result signal indicating that the number of mark elements is a specified number is received, if a detection result signal indicating that the number of mark elements is not a mark or a base color is received, it is determined that the mark is not a mark. The GC mark recognition process in the specific mark recognition chip 26 has been described as a software program here, but at least a part of it can be configured as a dedicated hardware circuit.
[0035]
FIG. 9 shows a flow of mark interval inspection (FIG. 8, S108) and mark number count (FIG. 8, S110) for image data. Here, the specific mark is detected within the range of the designated number of lines from the edge of the document. As described above, it is assumed that the specific marks are arranged at a predetermined interval. The input image data is sequentially scanned to detect a specific mark pattern. The coordinate position of the binary data is represented by (X, Y). First, the detection mark element number nCount is initialized to 0, and the Y coordinate value is initialized to 0 (S200). Next, the value of the X coordinate is initialized to 0 (S202). Next, a mark element candidate is detected by pattern matching using a mark element filter of 11 × 11 pixels (S204). If a mark element candidate is detected, it is next determined whether or not it is the first mark element (S206). If it is the first mark element, the number nCount of detected mark elements is incremented and a flag PreFlag indicating that a mark element candidate has been detected is set (S208). Then, the X coordinate of the center of the detected mark element is set to x1, and the current X value is set to the previously detected mark element position PreX (S222). If the mark element candidate is not the first mark element (NO in S206), the X coordinate of the mark element is set to x2. Next, it is determined whether or not the distance (absolute value of the difference between x2 and x1) between the mark element and the previous detection mark element position PreX is a natural number multiple of the specified interval W (S210). If it is a natural number multiple of the specified interval W (that is, if abs (x2-x1) = nW ± 1), the detection mark element number nCount is incremented (S212). ± 1 indicates an allowable range. If the flag PreFlag is 0 (YES in S214), the number of detected mark elements nCount is incremented (S216), and a mark element candidate previously determined not to be a mark element candidate is also handled as a detected mark element. The flag PreFlag is set to 1 (S218). Then, the value of the current X coordinate is set to the previously detected mark element position PreX (S222). If it is not the specified interval W (NO in S210), the flag PreFlag is set to 0 (S220). Then, the current value of X is set to the previous detection mark element position PreX (S222).
[0036]
Next, X is incremented to move the position in the main scanning direction (S224). If X is within the X detection range (NO in S226), the process returns to step S204 to continue the mark element detection.
[0037]
When X is not within the X detection range (YES in S226), Y is incremented and the position is moved in the sub-scanning direction (S228). If Y is within the Y detection range (NO in S230), the process returns to step S202 to continue the mark element detection. When Y is not within the Y detection range (YES in S230), the process ends. The value of nCount at this time is the number of detected marks.
[0038]
FIG. 10 shows an example of the mark number counting operation. In this example, mark element candidates indicated by rectangles are detected in the order shown in the figure. When the first mark element candidate is detected, the detected element number nCount becomes 1, and the flag PreFlag is set to 1. Next, a second mark element candidate is detected, but it is not a mark element because it does not exist at a specified interval position. Therefore, the flag PreFlag is set to 0. Next, the third mark element candidate is detected, but is not located at a specified interval, so it is determined that the element is not an element, and the flag PreFlag is set to 0. When the fourth mark element candidate is detected, it exists at the specified interval, so the number of detected elements nCount is 2, and it is determined that the previous element candidate is not at the specified interval. As a result, the number n of detected elements is increased to 3. The flag PreFlag is set to 1.
[0039]
【The invention's effect】
A color and shape similar to a mark are mixed in a general image, and even if a specific mark is detected by the mark detection unit, if the mark color ratio or background color ratio does not satisfy the conditions, it is not recognized as a mark. It is possible to prevent a general image from being erroneously recognized as a mark.
Further, since the accuracy with respect to erroneous recognition is increased, the mark itself may not have a complicated shape that is difficult to be erroneously recognized. Therefore, the configuration of the mark detection unit can be simplified.
[Brief description of the drawings]
FIG. 1 is a flowchart showing first generation copy processing;
[Figure 2] Diagram of an example of a generation copy
FIG. 3 is a flowchart showing a second generation copy process.
FIG. 4 is a diagram showing lines necessary for recognition
FIG. 5 is a diagram showing detection processing when the originals are placed differently.
FIG. 6 is a block diagram showing the overall configuration of a color copying machine.
FIG. 7 is a diagram for explaining the concept of mark recognition
FIG. 8 is a flowchart of specific mark recognition in the specific mark recognition chip.
FIG. 9 is a flowchart of mark interval inspection and mark count counting for image data.
FIG. 10 is a diagram for explaining a mark number counting operation;
[Explanation of symbols]
10 CCD sensor, 22 GC mark removal unit, 24 specific document recognition circuit, 26 specific mark recognition chip, 28 sharpness correction unit, 30 HVC adjustment unit, 32 color space conversion unit, 34 additional processing unit, 100 CPU, 102 ROM, 106 Processing circuit.

Claims (5)

A discriminating means for discriminating a specific mark of a predetermined color and shape;
First counting means for counting the number of specific marks determined by the determining means;
Second counting means for counting the number of pixels of the color of the specific mark;
Determining means for validating discrimination by the discriminating means when the count result of the first counting means is equal to or greater than the first threshold value and the count result of the second count means is equal to or less than the second threshold value; An image processing apparatus comprising: further,
A third counting means for counting the number of pixels of the background color of the specific mark;
The image processing apparatus according to claim 1, wherein the determination unit validates the determination by the determination unit when a count result of the third count unit is equal to or greater than a third threshold value. at least,
The count result of the first counting means is not more than a fourth threshold value different from the first threshold value,
The count result of the second count means is not more than a fifth threshold value different from the second threshold value,
The determination by the determination unit is invalidated when the count result of the third count unit satisfies any one of the sixth threshold value different from the third threshold value. Image processing device. A discriminating means for discriminating a specific mark from the pattern shape;
First counting means for counting the number of specific marks determined by the determining means;
Third counting means for counting the number of pixels of the background color of the specific mark;
An image processing apparatus comprising: a determination unit that determines that the determination by the determination unit is valid when the count result of the third count unit is equal to or greater than a third threshold value. The determination of the specific mark by the determination unit and the count of the number of pixels of the color of the specific mark by the second counting unit are performed in parallel with the extraction of the pixel of the color of the specific mark from the input image by the color extraction unit. The image processing apparatus according to claim 1, wherein the image processing apparatus is executed.

Images