JP4382472B2 - Halftone dot area identification device and halftone dot area identification method - Google Patents

Description

  The present invention relates to a halftone dot region identification device and a halftone dot region identification method that automatically identify a halftone dot region in image data.

  2. Description of the Related Art Conventionally, in a copying machine, image area separation processing is performed on an image input by a scanner, character areas and halftone dot areas in the image are identified, and image processing suitable for each area is performed. As a technique for identifying a halftone dot region, a method is often used in which a peak pixel that is the central pixel of a halftone dot is detected and the halftone dot region is identified from the density and period of the peak pixel.

  FIG. 28 is a diagram for explaining a halftone image based on peak points. FIG. 29 and FIG. 30 are diagrams for explaining a halftone image by cross points. The peak point is a central pixel in a square area surrounded by a thick line in FIG. 28, and can be detected by a density difference between the central pixel and the peripheral pixels.

  The apparatus described in Patent Document 1 is a method for identifying a halftone dot region using features of halftone dots other than peak pixels, and detects a cross point instead of a peak point. The cross point is the central pixel of the square area surrounded by the thick line in FIG. 30 is an enlarged view of the cross point shown in FIG.

  In this document, a pixel having a maximum value in one direction on the diagonal line and a minimum value in the other direction is detected as a cross point. Referring to FIG. 30, the density takes a maximum value on the diagonal line going from the upper left to the lower right through the central pixel, and the density takes the minimum value on the diagonal line going from the upper right to the lower left. The cross point has the advantages that the reference area size can be set smaller than the peak point and that erroneous separation is less likely to occur at the character portion. This is because a reading point, a turbid point, or the like is easily detected as a peak point at the time of detecting a peak point, but there is no explicit pattern at the cross point that is easily detected.

JP 2002-142109 A

  However, in the apparatus described in this document, it is mentioned that peak point detection is used when it is easier to detect highlight halftone dots rather than cross points. However, in the cross point detection method described above, the 40% halftone dot that is slightly away from 50% does not contact the halftone dots, and there are white pixels between the halftone dots, so there is no maximum value on the diagonal line. Conversely, there is no minimum value at 60% halftone dots. In that case, 40% halftone dots and 60% halftone dots must be covered by peak point detection. For this reason, since the range covered by the cross point detection is reduced, the size of the reference area by the cross point detection and the effect of suppressing the character part erroneous separation are extremely reduced. In fact, it can be said that there is almost no state. Therefore, in the case of covering with peak point detection, the meaning of performing cross point detection is lost. Also, 40% halftone dots with a low number of lines per 65 lines and the reading marks existing in the character portion are about the same size if the characters are about 8 points, and it is difficult to identify by peak point detection. There is a problem that the use of peak point detection is limited.

  Patent Document 1 does not mention the case of color halftone dots. However, if the judgment threshold at the time of peak point detection or cross point detection is tightened, halftone dots overlap each other in a color halftone document. When it is difficult to see (a halftone dot does not exist in isolation), there is a problem that a peak point or a cross point is not detected.

  The most prominent case is a 3C gray halftone dot represented by three color dots of C, M, and Y (see FIG. 17). Since the halftone dot overlaps, a unique pattern of Rosetta is created. There has been a problem that it is very difficult to detect peak points and cross points.

  The present invention has been made in view of the above, and an object of the present invention is to provide a cross point detection method capable of covering 50% halftone dots and 60% halftone dots by cross point detection. It is an object of the present invention to provide a high-accuracy halftone dot region identification device and a halftone dot region identification method that take advantage of the cross point detection in which erroneous separation is unlikely to occur.

  Another object of the present invention is to decompose an input color RGB signal into a CMY version of an original and detect a peak point and a cross point for each of the plates, thereby reducing the peak point and the 3D gray dot without relaxing the determination threshold. A high-accuracy halftone dot area identification device that can detect crosspoints and enables detection of crosspoints even in the case of mixed-color color halftone dots, reducing the occurrence of false separation in the character area, and halftone dot area discrimination Is to realize the method.

In order to solve the above-described problems and achieve the object, the invention according to claim 1 is a halftone dot region identification device, which refers to a region in image data and rises in the first direction in the image data. A first edge detecting means for detecting an edge and a falling edge; and a first edge detecting means for detecting a rising edge and a falling edge in a second direction orthogonal to the first direction in the image data with reference to the region. The edge detected by the second edge detecting means and the edges detected by the first and second edge detecting means are referred to in a predetermined area, and the sum of the rising edges in the first direction in the referred predetermined area Calculate the sum of the falling edges in the first direction, the sum of the rising edges in the second direction, and the sum of the falling edges in the second direction , Unified code of each value computed, the minimum value among the values obtained by unifying the code is equal to or greater than or equal to the threshold value for determining a deviation of an edge pixel having a dot feature based on the determination result And a halftone dot feature detecting means for detecting.

According to the first aspect of the present invention, in the halftone dot region identification device, the first edge detecting means refers to the region in the image data, and the rising edge and the falling edge in the first direction in the image data. Detect edges. The second edge detecting means refers to the region and detects a rising edge and a falling edge in a second direction orthogonal to the first direction in the image data. The halftone dot area detecting means refers to the edges detected by the first and second edge detecting means in a predetermined area, the sum of rising edges in the first direction in the referred predetermined area, Calculate the sum of the falling edges in the direction, the sum of the rising edges in the second direction, and the sum of the falling edges in the second direction, and unify the sign of each calculated value. It is determined whether or not the minimum value is equal to or greater than a threshold value for determining the edge bias, and pixels having halftone dot characteristics are detected based on the determination result . With this configuration, since halftone pixels are detected by using the symmetry of the rising and falling edges in the two directions , whether or not each edge exists in a biased manner , the halftone dot region can be made highly accurate. Can be identified.

  According to a second aspect of the present invention, in the halftone dot region identification device according to the first aspect, the halftone dot feature detection means is a center pixel between nearest halftone dots as the pixel having the halftone dot feature. It is characterized by detecting a certain cross point pixel.

  According to the second aspect of the present invention, the halftone dot feature detection means detects a cross dot pixel that is a central pixel between the closest halftone dots as a pixel having a halftone dot feature. With this configuration, since the cross-point pixel is detected using the symmetry of the rising edge and the falling edge in the first direction and the symmetry of the rising edge and the falling edge in the second direction, Compared with halftone dot region identification by pixel detection alone and halftone dot region identification by cross point detection using density minimum and / or maximum values, the device can be scaled down and erroneous detection of character parts can be suppressed. The dot area can be identified with high accuracy.

  According to a third aspect of the present invention, in the halftone dot region identification device according to the first or second aspect, each of the first edge detection unit and the second edge detection unit uses a first-order differential filter to perform edge detection. It is a thing which detects this.

  According to the third aspect of the invention, each of the first edge detection means and the second edge detection means detects the edge using the first-order differential filter. With this configuration, since the edges in the first direction and the second direction are detected using the first-order differential filter, there are cases where the first direction and the second direction are not strictly detected, or halftone dots are not clear. However, it can be generally handled, and it is difficult to cause large misidentification, so that the halftone dot region can be identified with high accuracy.

  According to a fourth aspect of the present invention, in the halftone dot region identifying device according to any one of the first to third aspects, the halftone dot characteristic detecting means is an edge detected by the first edge detecting means. When the data and the edge data detected by the second edge detecting means are compared and it is determined that the absolute values of the indices representing the sizes of both edges are substantially equal, it is determined that the pixel is not a cross-point pixel. It is characterized by.

  According to the fourth aspect of the present invention, the halftone dot characteristic detecting means compares the edge data detected by the first edge detecting means and the edge data detected by the second edge detecting means, and both edges are detected. When it is determined that the absolute values of the indices representing the sizes of the pixels are substantially equal, it is determined that the pixel is not a cross point pixel. With this configuration, in the first edge and second edge detection unit results, when the response is almost zero in the middle direction between the first direction and the second direction, it is difficult to distinguish the pixel from the cross-point pixel. A fine line image having a directional angle can be accurately excluded from halftone dot feature pixels, and erroneous identification at a character portion can be suppressed, so that a halftone dot region can be identified with high accuracy.

According to a fifth aspect of the present invention, in the halftone dot area identifying device according to any one of the first to fourth aspects, the halftone dot feature detecting means refers to a plurality of predetermined reference areas having different sizes. And a sum of rising edges in the first direction, a sum of falling edges in the first direction, a sum of rising edges in the second direction, and the second for each of the plurality of predetermined reference regions . Calculate the sum of the falling edges of the direction, unify the sign of each calculated value, determine whether the minimum value among the values that unified the sign is greater than or equal to the threshold, and based on the determination result , It is characterized by detecting pixels having a halftone dot feature.

According to the invention of claim 5, the halftone dot feature detecting means refers to a plurality of predetermined reference areas having different sizes , and sums the rising edges in the first direction for each of the plurality of predetermined reference areas . Calculate the sum of the falling edges in the first direction, the sum of the rising edges in the second direction, and the sum of the falling edges in the second direction, unify the sign of each calculated value, and unify the sign It is determined whether or not the minimum value among the values is equal to or greater than the threshold value, and pixels having halftone dot characteristics are detected based on the determination result . With this configuration, the halftone dot feature detection means refers to a plurality of regions of different sizes and detects cross point pixels using the symmetry of the rising edge and the falling edge for each reference region. A halftone dot region can be identified with high accuracy regardless of the number of lines from a point to a high line number halftone dot.

According to a sixth aspect of the present invention, in the halftone dot region identification device according to any one of the first to fifth aspects, the third direction is an intermediate direction between the first direction and the second direction. Third edge detection means for detecting a rising edge and a falling edge in the direction of the fourth direction; fourth edge detection means for detecting a rising edge and a falling edge in a fourth direction orthogonal to the third direction; The halftone dot feature detecting means refers to the edge detected by the third and fourth edge detecting means in the predetermined area, and the third direction in the referred predetermined area The sum of the rising edges, the sum of the falling edges in the third direction, the sum of the rising edges in the fourth direction, and the sum of the falling edges in the fourth direction were calculated. Unifying the sign of the value, determining whether or not the minimum value in the unified value of the sign is greater than or equal to the threshold value, the determination result using the edges in the first direction and the second direction, A pixel having a halftone dot feature is detected based on one of determination results using edges in the third direction and the fourth direction .

According to the sixth aspect of the present invention, the third edge detecting means for detecting the rising edge and the falling edge in the third direction, which is an intermediate direction between the first direction and the second direction, And a fourth edge detecting means for detecting a rising edge and a falling edge in a fourth direction orthogonal to the direction of 3. The halftone dot characteristic detecting means is detected by the third and fourth edge detecting means. The edge in the predetermined area, the sum of the rising edges in the third direction, the sum of the falling edges in the third direction, the sum of the rising edges in the fourth direction, And the sum of the falling edges in the fourth direction, the signs of the calculated values are unified, it is determined whether or not the minimum value among the unified values of the signs is equal to or greater than a threshold, Direction and second Determination result using the edge of the counter, on the basis of one of the determination result using the third direction and the fourth direction of the edge, detecting a pixel having a dot feature. With this configuration, similarly to the first and second edge detection means, based on the symmetry between the rising edge and the falling edge in the third direction and the symmetry between the rising edge and the falling edge in the fourth direction. Thus, pixels having a halftone dot feature are detected, so that it is possible to cope with color halftone dots and document skew, and a halftone dot region can be identified with high accuracy even for color halftone images and skewed originals. .

According to a seventh aspect of the present invention, in the halftone dot region identification device according to the sixth aspect, the halftone dot feature detection means uses the determination result using the edges in the first direction and the second direction. Based on the determination result using the first halftone dot feature detection means for detecting the halftone dot feature pixel and the edges in the third direction and the fourth direction, the pixel having the halftone dot feature A second halftone dot feature detecting means for detecting the first halftone dot feature pixel, and a first correction means for correcting the halftone dot feature pixel data detected by the first halftone dot feature detection means, And second correction means for correcting halftone dot feature pixel data detected by the second halftone feature detection means.

According to the seventh aspect of the invention, the halftone dot feature detecting means detects the pixel having the halftone dot feature based on the determination result using the edges in the first direction and the second direction . Halftone dot feature detection means; and second halftone dot feature detection means for detecting pixels having halftone dot characteristics based on determination results using edges in the third direction and the fourth direction ; Further, a first correction unit for correcting the halftone dot feature pixel data detected by the first halftone dot feature detection unit and a first correction unit for correcting the halftone dot feature pixel data detected by the second halftone dot feature detection unit. Two correction means. With this configuration, since the halftone dot image detected by the halftone dot feature detection unit is corrected, the detected halftone dot region can be identified with high accuracy.

The invention according to claim 8 is a halftone dot region identification method, wherein a first edge for detecting a rising edge and a falling edge in a first direction in the image data with reference to a region in the image data. A detection step; a second edge detection step of detecting a rising edge and a falling edge in a second direction orthogonal to the first direction in the image data with reference to the region; and the first edge The edge data detected by the detection step and the second edge detection step is referred to in a predetermined region, and the sum of rising edges in the first direction in the referred to the predetermined region, the first direction Calculate the sum of the falling edges, the sum of the rising edges in the second direction, and the sum of the falling edges in the second direction, and calculate each value Unified code, and determine the constant step you determine whether the minimum value among the values obtained by unifying the code is equal to or higher than the threshold for determining the deviation of the edge, the network based on the determination result of the previous SL-size constant step And a halftone dot feature detecting step of detecting pixels having point features.

According to the invention of claim 8, there is provided a halftone dot region identification method, wherein the first edge detection step refers to the region in the image data, and the rising edge and the falling edge in the first direction in the image data. Is detected. In the second edge detection step, a rising edge and a falling edge in a second direction orthogonal to the first direction in the image data are detected with reference to the region. The determine a constant step, the sum of the first edge detecting step and second edge data detected by the edge detection step, the first direction of the rising edge in the reference in a predetermined area, referenced predetermined region Calculate the sum of the falling edges in the first direction, the sum of the rising edges in the second direction, and the sum of the falling edges in the second direction, unify the sign of each calculated value, and unify the sign It is determined whether or not the minimum value among the measured values is equal to or greater than a threshold value for determining edge bias . The dot feature detection step, detecting a pixel having a dot feature based on the determination result of determine a constant step. With this configuration, since halftone pixels are detected by using the symmetry of the rising and falling edges in the two directions , whether or not each edge exists in a biased manner , the halftone dot region can be made highly accurate. Can be identified.

  According to a ninth aspect of the present invention, in the halftone dot region identifying method according to the eighth aspect, the halftone dot feature detection step is performed by using a central pixel between nearest halftone dots as the pixel having the halftone dot feature. It is characterized by detecting a certain cross point pixel.

  According to the ninth aspect of the invention, the halftone dot feature detecting step detects a cross point pixel that is a central pixel between the closest halftone dots as a pixel having a halftone dot feature. With this configuration, since the cross-point pixel is detected using the symmetry of the rising edge and the falling edge in the first direction and the symmetry of the rising edge and the falling edge in the second direction, Compared with halftone dot region identification by pixel detection alone and halftone dot region identification by cross point detection using density minimum and / or maximum values, the device can be scaled down and erroneous detection of character parts can be suppressed. The dot area can be identified with high accuracy.

  The invention according to claim 10 is the halftone dot region identification method according to claim 8 or 9, wherein each of the first edge detection step and the second edge detection step uses a first-order differential filter. It is a thing which detects this.

  According to the tenth aspect of the present invention, the first edge detection step and the second edge detection step each detect an edge using a first-order differential filter. With this configuration, since the edges in the first direction and the second direction are detected using the first-order differential filter, there are cases where the first direction and the second direction are not strictly detected, or halftone dots are not clear. However, it can be generally handled, and it is difficult to cause large misidentification, so that the halftone dot region can be identified with high accuracy.

  The invention according to claim 11 is the halftone dot region identification method according to any one of claims 8 to 10, wherein the halftone dot feature detection step is an edge detected by the first edge detection step. When the data and the edge data detected by the second edge detection step are compared and it is determined that the absolute values of the indices indicating the sizes of both edges are substantially equal, it is determined that the pixel is not a cross-point pixel. It is characterized by.

  According to the invention of claim 11, the halftone dot feature detecting step compares the edge data detected by the first edge detecting step and the edge data detected by the second edge detecting step, and both edges When it is determined that the absolute value of the index indicating the size of the pixel is substantially equal, it is determined that the pixel is not a cross point pixel. With this configuration, in the first edge and second edge detection unit results, when the response is almost zero in the middle direction between the first direction and the second direction, it is difficult to distinguish the pixel from the cross-point pixel. A fine line image having a directional angle can be accurately excluded from halftone dot feature pixels, and erroneous identification at a character portion can be suppressed, so that a halftone dot region can be identified with high accuracy.

The invention according to claim 12, in the halftone area identification method according to any one of claims 8 to 11, before SL-size constant step, with reference to the plurality of predetermined reference region having different sizes , A sum of rising edges in the first direction, a sum of falling edges in the first direction, a sum of rising edges in the second direction, and the second direction for each of the plurality of predetermined reference regions Calculating the sum of the falling edges of each of the values, unifying the sign of each calculated value, determining whether or not the minimum value among the unified values of the signs is equal to or greater than the threshold value, A pixel having a halftone dot feature is detected based on a determination result determined for each of the plurality of predetermined reference regions.

According to the invention of the claim 12, determine the constant process refers to the plurality of predetermined reference region having different sizes, the sum of the first direction of a rising edge for each of the plurality of predetermined reference area, The sum of the falling edges in the first direction, the sum of the rising edges in the second direction, and the sum of the falling edges in the second direction are calculated, and the sign of each calculated value is unified. It is determined whether or not the minimum value among the unified values is equal to or greater than the threshold value, and the halftone dot feature detection step has a halftone dot feature based on a determination result determined for each of a plurality of predetermined reference areas. Detect a pixel. This configuration, in determine the constant step, with reference to the plurality of areas of different sizes, to determine the symmetry of the falling rising edge and every reference area edge, the determined symmetry in the halftone characteristic detection step Since the cross-point pixel is detected based on this, the halftone dot region can be identified with high accuracy regardless of the number of lines from the low line number halftone dot to the high line number halftone dot.

The invention according to claim 13 is the halftone dot region identification method according to any one of claims 8 to 12, further comprising a third direction that is an intermediate direction between the first direction and the second direction. A third edge detecting step for detecting a rising edge and a falling edge in the direction of the fourth direction; a fourth edge detecting step for detecting a rising edge and a falling edge in a fourth direction orthogonal to the third direction; wherein the pre-Symbol-size constant step, the edge data detected by said first edge detecting step and second edge detecting step, referred to in a given area, said at referenced predetermined region first Sum of rising edges in one direction, sum of falling edges in the first direction, sum of rising edges in the second direction, and sum of falling edges in the second direction And, calculating a unified code of each value, first a determination step minimum value is equal to or more than the threshold in the value of unified code, the third edge detection step and the The edge data detected by the edge detection step 4 is referred to in the predetermined area, and the sum of the rising edges in the third direction and the falling edge in the third direction in the referred to the predetermined area , The sum of the rising edges in the fourth direction, and the sum of the falling edges in the fourth direction, and unifying the sign of each calculated value, and the minimum value among the unified values of the signs There are those comprising a second determination step of determining whether or not more than the threshold value, the network characteristic point detection step, the determination of the determination result of the first determination process and the second determination step based on one of the results, halftone Characterized in that it is designed to detect the pixels having the symptoms.

According to the thirteenth aspect of the present invention, the third edge detecting step for detecting the rising edge and the falling edge in the third direction which is an intermediate direction between the first direction and the second direction, includes a fourth edge detection step of detecting the fourth direction of the rising and falling edges orthogonal to the third direction and a, determine the constant step, by the first edge detecting step and second edge detecting step The detected edge data is referred to in a predetermined area, and the sum of rising edges in the first direction, the sum of falling edges in the first direction, and the rising edge in the second direction in the referred predetermined area And the sum of the falling edges in the second direction are calculated, the signs of the calculated values are unified, and it is determined whether or not the minimum value among the unified values of the signs is equal to or greater than a threshold value . 1 The sum of the determination step and, a third edge detection step and the fourth edge data detected by the edge detection step, the third direction of the rising edge of the reference in a predetermined area, referenced predetermined region, Calculate the sum of the falling edges in the third direction, the sum of the rising edges in the fourth direction, and the sum of the falling edges in the fourth direction, unify the signs of the calculated values, and unify the signs And a second determination step of determining whether or not a minimum value among the values is equal to or greater than a threshold value . The halftone dot feature detection step includes the determination result of the first determination step and the second determination step. A pixel having a dot feature is detected based on one of the determination results of the process. With this configuration, similar to the first and second edge detection steps, based on the symmetry between the rising edge and the falling edge in the third direction and the symmetry between the rising edge and the falling edge in the fourth direction. Thus, pixels having a halftone dot feature are detected, so that it is possible to cope with color halftone dots and document skew, and a halftone dot region can be identified with high accuracy even for color halftone images and skewed originals. .

According to a fourteenth aspect of the present invention, in the halftone dot region identifying method according to the thirteenth aspect, the halftone dot feature detection step is a pixel having a halftone dot feature based on a determination result of the first determination step. A first halftone dot feature detection step for detecting a halftone dot feature, a second halftone dot feature detection step for detecting a pixel having a halftone dot feature based on a determination result of the second determination step , and the first halftone dot detection A first correction step for correcting halftone dot feature pixel data detected by the point feature detection step; and a second correction step for correcting halftone dot feature pixel data detected by the second halftone feature detection step; It is characterized by including.

According to the fourteenth aspect of the present invention, the halftone dot feature detecting step includes a first halftone dot feature detecting step of detecting pixels having halftone dot features based on the determination result of the first determining step, The second halftone dot feature detection step for detecting pixels having halftone dot features and the halftone dot feature pixel data detected by the first halftone dot feature detection step are corrected based on the determination result of the second determination step. A first correction step, and a second correction step of correcting the halftone dot feature pixel data detected by the second halftone dot feature detection step. With this configuration, since the halftone image detected by the halftone feature detection step is corrected, the detected halftone area can be identified with high accuracy.

Further, the invention relates to a halftone dot region identification apparatus according to claim 1, further input image data comprising a plurality of color signals, different predetermined plurality of said plurality of color signals and attributes according to claim 15 Color conversion means for converting to image data consisting of color signals, peak point detection means for detecting a peak point which is a central pixel of halftone dots from image data after color conversion processing by the color conversion means, and the input Halftone dot area identifying means for identifying a halftone dot area in the image data, wherein the first edge detecting means refers to the area in the image data after color conversion processing by the color converting means, and the image data The rising edge and the falling edge in the first direction are detected, and the second edge detecting means refers to the area in the image data after the color conversion process, and Detecting a rising edge and a falling edge in the second direction, and the halftone dot characteristic detecting means detects the halftone dot based on a determination result using the edges in the first direction and the second direction. A cross point that is a central pixel between nearest halftone dots is detected as a pixel having characteristics, and the halftone dot region identification unit is configured to detect a halftone dot region in the input image data based on the peak point and the cross point. It is characterized by identifying.

According to the fifteenth aspect of the present invention , the color conversion means further converts the input image data composed of a plurality of color signals into image data composed of a plurality of predetermined color signals having different attributes from the plurality of color signals. . The peak point detection means detects a peak point that is a central pixel of a halftone dot from the image data after the color conversion process. A halftone dot region identification unit identifies a halftone dot region in the input image data. The first edge detection unit detects a rising edge and a falling edge in the first direction in the image data with reference to an area in the image data after the color conversion processing by the color conversion unit. The second edge detecting means detects a rising edge and a falling edge in the second direction in the image data with reference to the area in the image data after the color conversion process. The halftone dot feature detection means detects a cross point that is a central pixel between the closest halftone dot dots as a pixel having a halftone dot feature based on the determination result using the edges in the first direction and the second direction. To do. The halftone dot region identification means identifies a halftone dot region in the input image data based on the peak point and the cross point . With this configuration, after color conversion of input image data composed of a plurality of color signals into image data composed of a plurality of color signals having attributes different from those of the plurality of color signals, a peak point and a cross point are detected and a halftone dot region is detected. In order to identify, halftone dot area by detecting peak point and cross point after converting into color space which is easy to detect even in mixed color color halftone dot which was difficult to detect in input color space Can be identified with high accuracy. In addition, for mixed color halftone dots, as in the case of single color halftone dots, it is possible to detect cross points that can reduce the scale of the device and reduce erroneous detection of character portions. it can.

  According to a sixteenth aspect of the present invention, in the halftone dot region identification device according to the fifteenth aspect, the color conversion means includes the plurality of input image data including a plurality of color signals and a signal corresponding to a process color. The image data is converted into image data including a plurality of predetermined color signals having attributes different from those of the color signals.

  According to the sixteenth aspect of the present invention, the color conversion means obtains input image data comprising a plurality of color signals from a plurality of color signals including attributes corresponding to process colors and a plurality of predetermined color signals having different attributes. Is converted into image data. With this configuration, since the input image data is converted into a signal corresponding to the process color and then the peak point and the cross point are detected, even in the case of the 3C color halftone dot, the apparatus can be reduced in size and the character portion as in the case of the single color halftone dot. Since cross point detection that can reduce erroneous detection is possible, a halftone dot region can be identified with high accuracy.

The invention according to claim 17, in the halftone region identification apparatus according to claim 15 or 16, wherein the peak point detecting means is adapted to detect the dots of at least low and high concentrations, the network The point feature detecting means is characterized by detecting at least a halftone dot having an intermediate density.

According to the seventeenth aspect of the present invention, the peak point detecting means detects at least low density and high density halftone dots, and the halftone characteristic detecting means detects at least intermediate density halftone dots. With this configuration, the peak point detection and the cross point detection need only be able to detect at least the halftone dot density with high detection efficiency, so the character portion generated when detecting the halftone dot density with low detection efficiency is detected. Since false detection can be reduced, a halftone dot region can be identified with high accuracy.

Further, the invention according to claim 18 is the halftone dot region identification device according to any one of claims 15 to 17, wherein the peak point detection means is the predetermined plurality of predetermined values converted by the color conversion means. The peak point is detected for each color signal, and the halftone dot feature detection unit detects the cross point for each of the predetermined plurality of color signals converted by the color conversion unit. It is characterized by that.

According to the eighteenth aspect of the present invention, the peak point detection means detects the peak point for each of a plurality of predetermined color signals converted by the color conversion means. The halftone dot feature detection means detects a cross point for each of a plurality of predetermined color signals converted by the color conversion means. With this configuration, peak points and cross points are detected for each of a plurality of color signals, so that a halftone dot region can be identified with high accuracy even for a mixed color halftone dot of any color.

Further, the invention according to claim 19 is the halftone dot region identification device according to any one of claims 15 to 18, further comprising at least one of the detected data of the peak point and the cross point. And a correction unit that performs correction for each of the plurality of converted colors, and the halftone dot region identification unit identifies a halftone dot region based on the data corrected by the correction unit. It is characterized by being.

According to the invention of the claim 19, further based on one or more of the detected data of the peak point and the cross point for each transformed multiple colors, a correction means for correcting, web The dot area identifying means identifies the halftone dot area based on the data corrected by the correcting means. With this configuration, correction is performed for each color in accordance with the appearance state of peak points and cross points, so that isolated peak points and cross points can be more accurately excluded, so that a halftone dot region can be identified with high accuracy. be able to.

  Further, the invention according to claim 20 is the halftone dot region identification device according to any one of claims 15 to 19, wherein the peak point detection means is a density of a target pixel and a surrounding pixel surrounding the target pixel. It is determined whether or not the difference is equal to or greater than a predetermined threshold value, and when it is determined that the difference is not equal to or greater than the threshold value, it is determined that the target pixel is not a peak point.

  According to the twentieth aspect of the present invention, the peak point detection means determines whether or not the density difference between the target pixel and the surrounding pixels surrounding the target pixel is equal to or greater than a predetermined threshold value, and not In this case, it is determined that the target pixel is not a peak point. With this configuration, the peak point detection condition is that the density difference between the target pixel and the surrounding pixels is greater than or equal to a certain value, so it is possible to prevent erroneous detection of the peak point at the end point of the character portion, etc. A halftone dot region can be identified with high accuracy.

Further, the invention relates to a halftone dot region identification method according to claim 8, further input image data comprising a plurality of color signals, different predetermined plurality of said plurality of color signals and attributes according to claim 21 a color conversion step of converting the image data composed of color signals from the image data after the color conversion process in the color conversion process, a peak point detection step of detecting a peak point which is the center pixel of the halftone dots, the entering A halftone dot region identifying step for identifying a halftone dot region in the force image data , wherein the first edge detecting step refers to the region in the image data after the color conversion processing by the color conversion step, and A rising edge and a falling edge in the first direction in the data are detected, and the second edge detecting step refers to the area in the image data after the color conversion process, and Detecting a rising edge and a falling edge in the second direction in the data, and the halftone dot feature detecting step uses pixels between closest halftone dots as pixels having a halftone dot feature based on a determination result of the determining step. A cross point that is a central pixel is detected, and the halftone dot region identifying step identifies a halftone dot region in the input image data based on the peak point and the cross point .

According to the invention of claim 21, there is provided a halftone dot region identification method, and the color conversion step further comprises converting a plurality of predetermined image data having attributes different from those of the plurality of color signals into the input image data. Convert to image data consisting of color signals. In the peak point detection step, a peak point that is a central pixel of a halftone dot is detected from the image data after the color conversion process. The halftone dot region identifying step identifies a halftone dot region in the input image data. In the first edge detection step, a rising edge and a falling edge in the first direction in the image data are detected with reference to a region in the image data after the color conversion processing in the color conversion step. In the second edge detection step, the rising edge and the falling edge in the second direction in the image data are detected with reference to the area in the image data after the color conversion process. The halftone dot feature detection step detects a cross point, which is a central pixel between closest halftone dots, as a pixel having a halftone dot feature based on the determination result of the determination step. The halftone dot region identifying step identifies a halftone dot region in the input image data based on the peak point and the cross point . With this configuration, after color conversion of input image data composed of a plurality of color signals into image data composed of a plurality of color signals having attributes different from those of the plurality of color signals, a peak point and a cross point are detected and a halftone dot region is detected. In order to identify, halftone dot area by detecting peak point and cross point after converting into color space which is easy to detect even in mixed color color halftone dot which was difficult to detect in input color space Can be identified with high accuracy. In addition, for mixed color halftone dots, as in the case of single color halftone dots, it is possible to detect cross points that can reduce the scale of the device and reduce erroneous detection of character portions. it can.

  According to a twenty-second aspect of the present invention, in the halftone dot region identifying method according to the twenty-first aspect, the color conversion step includes the plurality of input image data including a plurality of color signals and a signal corresponding to a process color. The image data is converted into image data including a plurality of predetermined color signals having attributes different from those of the color signals.

  According to the twenty-second aspect of the present invention, in the color conversion step, input image data composed of a plurality of color signals is obtained from a plurality of color signals including attributes corresponding to process colors and a plurality of predetermined color signals having different attributes. Is converted into image data. With this configuration, since the input image data is converted into a signal corresponding to the process color and then the peak point and the cross point are detected, even in the case of the 3C color halftone dot, the apparatus can be reduced in size and the character portion as in the single color halftone dot. Since cross point detection that can reduce erroneous detection is possible, a halftone dot region can be identified with high accuracy.

The invention according to claim 23, in the halftone area identification method according to claim 21 or 22, wherein the peak point detection step is for detecting the dots of at least low and high concentrations, the network The point feature detection step is characterized by detecting at least a halftone dot having an intermediate density.

According to the twenty-third aspect of the present invention, the peak point detecting step detects at least low density and high density halftone dots, and the halftone dot characteristic detecting step detects at least intermediate density halftone dots. With this configuration, the peak point detection and the cross point detection need only be able to detect at least the halftone dot density with high detection efficiency, so the character portion generated when detecting the halftone dot density with low detection efficiency is detected. Since false detection can be reduced, a halftone dot region can be identified with high accuracy.

The invention according to claim 24 is the halftone dot region identification method according to any one of claims 21 to 23, wherein the peak point detection step includes a plurality of the predetermined plurality of points converted by the color conversion step. The peak point is detected for each color signal, and the halftone feature detection step detects the cross point for each of the predetermined plurality of color signals converted by the color conversion step. It is characterized by that.

According to the twenty-fourth aspect of the invention, the peak point detection step detects a peak point for each of a plurality of predetermined color signals converted by the color conversion step, and the halftone dot feature detection step A cross point is detected for each of a plurality of predetermined color signals converted by the conversion process. With this configuration, peak points and cross points are detected for each of a plurality of color signals, so that a halftone dot region can be identified with high accuracy even for a mixed color halftone dot of any color.

The invention according to claim 25 is the halftone dot region identification method according to any one of claims 21 to 24, and further includes at least one of the detected data of the peak point and the cross point. A correction step for correcting the plurality of colors converted in the color conversion step, and the halftone dot region identification step is based on the data subjected to the correction processing in the correction step. This is characterized by identifying a halftone dot region.

According to the twenty-fifth aspect of the present invention, the correction step further performs correction for each of the plurality of colors converted in the color conversion step based on the detected data of one or more of the peak point and the cross point. And the halftone dot region identifying step identifies the halftone dot region based on the data subjected to the correction processing in the correction step. With this configuration, correction is performed for each color in accordance with the appearance state of peak points and cross points, so that isolated peak points and cross points can be more accurately excluded, so that a halftone dot region can be identified with high accuracy. be able to.

  The invention according to claim 26 is the halftone dot region identification method according to any one of claims 21 to 25, wherein the peak point detection step includes the density of the target pixel and the surrounding pixels surrounding the target pixel. It is determined whether or not the difference is equal to or greater than a predetermined threshold value. When it is determined that the difference is not equal to or greater than the threshold value, the target pixel is determined not to be a peak point.

  According to the twenty-sixth aspect of the present invention, in the peak point detection step, it is determined whether or not the density difference between the target pixel and the surrounding pixels surrounding the target pixel is equal to or greater than a predetermined threshold value. In this case, it is determined that the target pixel is not a peak point. With this configuration, the peak point detection condition is that the density difference between the target pixel and the surrounding pixels is greater than or equal to a certain value, so it is possible to prevent erroneous detection of the peak point at the end point of the character portion, etc. A halftone dot region can be identified with high accuracy.

A halftone dot region identification device according to the present invention (Claim 1) uses a halftone dot pixel by utilizing whether or not each edge exists in a biased manner, that is, by using the symmetry of rising and falling edges in two directions. As a result, the halftone dot region can be identified with high accuracy.

  The halftone dot region identification device according to the present invention (Claim 2) has the symmetry between the rising edge and the falling edge in the first direction and the symmetry between the rising edge and the falling edge in the second direction. Compared with the conventional halftone dot region identification only by peak point pixel detection and the halftone dot region identification by cross point detection using density minimum and / or maximum values, Since the size can be reduced and erroneous detection of the character portion can be suppressed, the halftone dot region can be identified with high accuracy.

  In addition, since the halftone dot region identification device according to the present invention (Claim 3) detects the edges in the first direction and the second direction using the first-order differential filter, the first direction and the second direction are strictly limited. Even when the direction is not the direction or when the halftone dot is not clear, it can be dealt with in general, and it is difficult to cause large misidentification, so that the halftone dot region can be identified with high accuracy.

  In the halftone dot region identification device according to the present invention (claim 4), in the first edge and second edge detection unit results, the response is almost zero in the intermediate direction between the first direction and the second direction. In such a case, it is possible to accurately exclude the thin line image having an angle in the intermediate direction that is difficult to distinguish from the cross point pixel from the halftone dot feature pixel, and thus it is possible to suppress erroneous identification in the character portion. There is an effect that the halftone dot region can be identified with high accuracy.

  In the halftone dot area identification device according to the present invention (Claim 5), the halftone dot feature detection means refers to a plurality of areas having different sizes and determines the symmetry of the rising edge and the falling edge for each reference area. Since the cross-point pixel is detected by using this, there is an effect that the halftone dot region can be identified with high accuracy regardless of the number of lines from the low line number halftone dot to the high line number halftone dot.

  Further, the halftone dot region identification device according to the present invention (Claim 6) is similar to the first and second edge detection means, and the symmetry of the rising edge and the falling edge in the third direction, and the fourth Pixels with halftone dot characteristics are detected based on the symmetry of the rising edge and falling edge in the direction of, so that it is possible to cope with color halftone dots and document skew, and for color halftone images and skewed originals. In addition, the halftone dot region can be identified with high accuracy.

  In addition, since the halftone dot region identification device according to the present invention (Claim 7) corrects the halftone dot image detected by the halftone dot feature detection means, the detected halftone dot region can be identified with high accuracy. There is an effect.

Further, the halftone dot region identifying method according to the present invention (claim 8) is based on whether or not each edge exists in a biased manner, that is, using the symmetry of the rising and falling edges in the two directions. Since the dot pixel is detected, there is an effect that the halftone dot region can be identified with high accuracy.

  Also, the halftone dot region identification method according to the present invention (claim 9) has the symmetry between the rising edge and the falling edge in the first direction and the symmetry between the rising edge and the falling edge in the second direction. Compared with the conventional halftone dot region identification only by peak point pixel detection and the halftone dot region identification by cross point detection using density minimum and / or maximum values, Since the size can be reduced and erroneous detection of the character portion can be suppressed, the halftone dot region can be identified with high accuracy.

  In addition, since the halftone dot region identification method according to the present invention (Claim 10) detects the edges in the first direction and the second direction using the first-order differential filter, the first direction and the second direction are strictly limited. Even when the direction is not the direction or when the halftone dot is not clear, it can be dealt with in general, and it is difficult to cause large misidentification, so that the halftone dot region can be identified with high accuracy.

  Further, according to the halftone dot region identification method according to the present invention (claim 11), in the results of the first edge and the second edge detection unit, the response is almost zero in the intermediate direction between the first direction and the second direction. In such a case, it is possible to accurately exclude the thin line image having an angle in the intermediate direction that is difficult to distinguish from the cross point pixel from the halftone dot feature pixel, and thus it is possible to suppress erroneous identification in the character portion. There is an effect that the halftone dot region can be identified with high accuracy.

  In the halftone dot region identification method according to the present invention (Claim 12), in the symmetry determination step, the symmetry of the rising edge and the falling edge is determined for each reference region with reference to a plurality of regions having different sizes. Since the cross-point pixel is detected based on the symmetry determined in the halftone feature detection step, the halftone dot region is increased regardless of the number of lines from low to high halftone dots. There is an effect that it can be accurately identified.

  Further, in the halftone dot region identification method according to the present invention (claim 13), as in the first and second edge detection steps, the symmetry of the rising edge and the falling edge in the third direction, and the fourth Pixels with halftone dot characteristics are detected based on the symmetry of the rising edge and falling edge in the direction of, so that it is possible to cope with color halftone dots and document skew, and for color halftone images and skewed originals. In addition, the halftone dot region can be identified with high accuracy.

  In addition, since the halftone dot region identification method according to the present invention (claim 14) corrects the halftone dot image detected by the halftone dot feature detection step, the detected halftone dot region can be identified with high accuracy. There is an effect.

  A halftone dot region identification device according to the present invention (Claim 15) performs color conversion of input image data composed of a plurality of color signals into image data composed of a plurality of color signals having attributes different from those of the plurality of color signals. In order to identify the halftone dot area by detecting the peak point and the cross point, the input color space is converted into a color space that is easy to detect even in the mixed color halftone point, which is difficult to detect the peak point or the cross point. By detecting the peak point and the cross point later, the halftone dot region can be identified with high accuracy. In addition, for mixed color halftone dots, as in the case of single color halftone dots, it is possible to detect cross points that can reduce the scale of the device and reduce erroneous detection of character portions. There is an effect that can be done.

  In addition, since the halftone dot region identification device according to the present invention (Claim 16) detects the peak point and the cross point after converting the input image data into a signal corresponding to the process color, even if it is a 3C color halftone dot. As in the case of a single color halftone dot, it is possible to detect a cross point that can reduce the scale of the apparatus and reduce erroneous detection of a character portion.

  Further, the halftone dot region identification device according to the present invention (Claim 17) is only required to detect a halftone dot density with high detection efficiency at the peak point detection and the cross point detection. Since it is possible to reduce the erroneous detection of the character portion that has occurred when detecting the dot density, the halftone dot region can be identified with high accuracy.

  In addition, since the halftone dot area identifying device according to the present invention (claim 18) detects a peak point and a cross point for each of a plurality of color signals, a halftone dot area can be obtained even for a mixed color halftone dot of any color. It is possible to identify with high accuracy.

  In addition, since the halftone dot region identification device according to the present invention (Claim 19) performs correction for each color according to the appearance state of the peak point and the cross point, the peak point and the cross point existing in isolation are more accurately detected. Therefore, the halftone dot region can be identified with high accuracy.

  Further, in the halftone dot region identification device according to the present invention (claim 20), the peak point detection condition is that the density difference between the target pixel and the surrounding pixels is greater than or equal to a certain value. Since it is possible to prevent erroneous detection of the peak point, the halftone dot region can be identified with high accuracy.

  According to the halftone dot region identification method of the present invention (claim 21), after color conversion, input image data composed of a plurality of color signals is converted into image data composed of a plurality of color signals having attributes different from those of the plurality of color signals. In order to identify the halftone dot area by detecting the peak point and the cross point, the input color space is converted into a color space that is easy to detect even in the mixed color halftone point, which is difficult to detect the peak point or the cross point. By detecting the peak point and the cross point later, the halftone dot region can be identified with high accuracy. In addition, for mixed color halftone dots, as in the case of single color halftone dots, it is possible to detect cross points that can reduce the scale of the device and reduce erroneous detection of character portions. There is an effect that can be done.

  Further, in the halftone dot region identification method according to the present invention (claim 22), since the input image data is converted into a signal corresponding to the process color and the peak point and the cross point are detected, even if it is a 3C color halftone dot. As in the case of a single color halftone dot, it is possible to detect a cross point that can reduce the scale of the apparatus and reduce erroneous detection of a character portion.

  Further, the halftone dot region identification method according to the present invention (Claim 23) is only required to detect a halftone dot density with high detection efficiency for peak point detection and cross point detection. Since it is possible to reduce the erroneous detection of the character portion that has occurred when detecting the dot density, the halftone dot region can be identified with high accuracy.

  In addition, since the halftone dot region identifying method according to the present invention (claim 24) detects the peak point and the cross point for each of a plurality of color signals, the halftone dot region can be applied to a mixed color halftone dot of any color. It is possible to identify with high accuracy.

  In addition, since the halftone dot region identification method according to the present invention (claim 25) performs correction for each color according to the appearance state of the peak point and the cross point, the peak point and the cross point existing in isolation are more accurately detected. Therefore, the halftone dot region can be identified with high accuracy.

  In the halftone dot region identification method according to the present invention (claim 26), the peak point detection condition is that the density difference between the target pixel and the surrounding pixels is greater than or equal to a certain value. Since it is possible to prevent erroneous detection of the peak point, the halftone dot region can be identified with high accuracy.

(1. Embodiment)
(1.1. Overall configuration and detection procedure of halftone dot region detection device)
FIG. 1 is a functional block diagram of a halftone dot region detection apparatus according to the first embodiment. The halftone dot region detection apparatus 10 includes a first edge detection unit 1, a second edge detection unit 2, a halftone dot feature detection unit 3, a density correction unit 4, and an expansion unit 5.

  FIG. 2 is a functional block diagram of the first edge detector 1 and the second edge detector 2 shown in FIG. FIG. 3 is a flowchart for explaining the dot area detecting operation according to the first embodiment. FIG. 4 is a schematic diagram for explaining an edge portion in an image. Each function of the halftone dot region detection device will be described in the order of detection operation with reference to FIGS.

  A grid 403 indicated by reference numeral 503 in the region 501 in FIG. 4 is a cross point. The vertical edge corresponds to a portion where arrows A and B cross in the figure, and the + edge is A and the − edge is B. The horizontal edge corresponds to the part where arrows C and D cross in the figure, the + edge is C, and the − edge is D. The horizontal edge corresponds to the edge of the portion where the arrows C and D exceed, as illustrated in FIG. The + edge (= rising edge) shown in FIG. 30 is C, and the − edge (falling edge) is D.

  The first edge detection unit 1 detects a lateral edge. Similarly, the second edge detection unit 2 detects vertical edges.

  FIG. 5 is a schematic diagram illustrating an example of an edge detection filter.

  The halftone dot feature detection unit 3 utilizes the fact that the horizontal edge + edge C and −edge D are symmetrical as shown in FIG. 30, and the vertical edge + edge A and −edge B are vertically symmetrical. To detect a cross point.

  The density correction unit 4 refers to the density of the cross points and eliminates excessively detected cross points that cannot be halftone dot regions such as cross points that exist in isolation. Here, not only the density but also the cycle of the cross points may be detected and determined whether or not to be eliminated.

  The expansion unit 5 identifies the halftone dot region in a form including the halftone dot region and the edge of the halftone dot region by expanding the cross point after density correction.

  Further, the operation for executing the function of each unit will be described along the procedure of the halftone dot region detection operation shown in the flowchart of FIG. The first edge detection unit 1 and the second edge detection unit 2 include a horizontal edge detection filter 11, a vertical edge detection filter 21, sign determination units 12 and 22, absolute value conversion units 15 and 25, a subtractor 16, An absolute value conversion unit 17, a comparator 18, multipliers 13 and 23, and selectors 14 and 24 are included.

  The horizontal edge detection filter 11 detects the horizontal edge degree S of the local region using the 3 × 3 size horizontal edge detection filter 501 shown in FIG. The vertical edge detection filter 21 detects the vertical edge degree T of the local region using the 3 × 3 size vertical edge detection filter 502 shown in FIG. 5 (step S101).

  The absolute value conversion unit 15 and the absolute value conversion unit 25 calculate absolute values | S | and | T | of the horizontal edge degree S and the vertical edge degree T. The subtractor 16 calculates the difference | S | − | T | (step S102). The absolute value conversion unit 17 calculates the absolute value || S | − | T || of the difference | S | − | T | calculated by the subtractor 16.

  Here, the absolute value || S | − | T || of the difference calculated by the absolute value conversion unit 17 is added with the sign of + edge or −edge to be the value of the horizontal edge or the vertical edge. .

  The code determination unit 12 and the code determination unit 22 output 1 if the codes of S and T are +, and -1 if-. Using the output values output by the sign determination units 12 and 22, the multiplier 13 and the multiplier 23 perform multiplication on || S |-| T ||. || S |-| T || is signed by multiplication by multipliers 13 and 14, resulting in a + edge or a-edge.

  The comparator 18 compares the values of | S | and | T |, and if | S | is larger, sets the vertical edge to 0 and outputs it to the selector 24. If | T | is larger, the comparator 18 sets the horizontal edge to 0 and outputs it to the selector 14 (step S103).

  Here, the difference value between | S | and | T |, not the values of S and T, is used for the horizontal edge and the vertical edge. This is to make it almost zero. For example, there is a case where it is desired to perform cross inspection at the center pixel when the halftone dots in FIG. 4 are slightly separated from each other, but not to detect as a cross point in the gap between the diagonal lines in FIG. Since the values of S and T have a certain level of response even to diagonal edges, when the values of S and T are used directly as horizontal and vertical edge values, cross areas 401 and 402 in FIG. It may be impossible to clearly identify the region 450. Therefore, by taking the difference value, in the case of the hatched area 450, the vertical edge and the horizontal edge can be prevented from responding.

  In the configuration shown in FIG. 2, the first edge detection unit 1 and the second edge detection unit 2 share some circuits in common. Is also possible.

(1.2. Halftone feature detection unit)
FIG. 6 is a functional block diagram of the halftone dot feature detection unit shown in FIG. The halftone dot feature detection unit 3 includes mask 1 (31) to mask 4 (34), sign inversion units 35 and 36, sign inversion unit 39, minimum value selection units 37 and 40, threshold value determination units 38 and 41, and An adder 42 is provided.

  FIG. 7 is a schematic diagram showing an example of a mask used in the halftone dot feature detection unit of FIG. The mask 1 (71) is a mask for calculating the sum of the horizontal edges C (FIGS. 30 and 4), and a 5 × 5 size is used here. Similarly, the mask 2 is a mask that calculates the sum of the horizontal edge D, the mask 3 is the vertical edge A, and the mask 4 is the vertical edge B. The sum of the horizontal and vertical edges is calculated using these masks 1 to 4 (step S104).

  The sign inversion unit 35 and the sign inversion unit 36 invert the signs of the processing data from the mask 1 and the mask 3.

  The minimum value selection unit 37 selects the minimum value from the inverted detection data of the mask 1, the detection data of the mask 2, the detection data of the inverted mask 3, and the detection data of the mask 4.

  FIG. 8 is a diagram for explaining a pattern of cross points. The threshold determination unit 38 determines that the crossing point is 1 when the selected minimum value is equal to or greater than a predetermined threshold c_th (> 0) (step S105). What is detected here is a cross point such as “cross point 1” 81 in FIG. 8, where the horizontal edge C is −, D is +, the vertical edge A is −, and B is +. is there.

  The threshold value is judged using the minimum value if there is more than a predetermined value in the area where the + edge or − edge of the horizontal edge or vertical edge should exist (= a symmetric area). Therefore, it is determined whether or not each edge is biased. There are several other methods that can be substituted by the determination method using symmetry.

  In order to detect the pattern 82 of the cross point 2 shown in FIG. 8 in the same manner, the sign inversion unit 39 (FIG. 6) is used to invert all the signs and the results of the masks 1 to 4 are used. Then, it is detected by a method using the same threshold as that of the cross point 1 (step S106). Then, a logical OR of the cross point 1 and the cross point 2 is calculated by the adder 42, and is determined to be a cross point (step S107).

(1.3. Effect)
As described above, according to the halftone dot region identification device of the first embodiment, crossing is performed using the + edge and −edge symmetry of the horizontal edge and the + edge and −edge symmetry of the vertical edge. Since the dot is detected and the halftone dot region is identified by detecting the cross point, the cross point can be detected including a halftone dot slightly apart from 50% as shown in FIG.

  Further, since the horizontal edge and the vertical edge are detected using a two-dimensional primary differential filter, the horizontal edge and the vertical edge can be detected even if they are not strictly 0 degrees or 90 degrees. Further, even when a halftone dot such as a poor quality printed matter or a color halftone dot having a low density is not clear, it can be detected widely, and thus misidentification leading to a large image deterioration can be reduced.

  Further, since the response is almost zero at an oblique edge that is an edge in the 45 ° direction, it can be clearly distinguished from the oblique line 450 in FIG. 4, and misidentification at the character portion can be reduced.

  FIG. 9 is a schematic diagram for explaining peak points. The above-described detection operation has been described with respect to the configuration in which the cross point detection unit is provided to detect the cross point. However, with respect to the peak point illustrated in FIG. 9, each of the horizontal edge and vertical edge + edge and − It is possible to detect a peak point using the symmetry of the edge. This can be realized by changing the combination of the masks shown in FIG. 7 in accordance with the pattern of FIG. For example, if masks 3 and 4 are used for horizontal edges C and D and masks 1 and 2 are used for vertical edges A and B, peak points can be detected.

(1.4. Modification 1)
FIG. 10 is a functional block diagram of Modification 1 of the halftone dot feature detection unit.

  The halftone dot feature detection unit 3 according to the first embodiment refers to the detection result of the horizontal edge and the vertical edge in the 5 × 5 area, but in the case of 5 × 5, the dot feature detection unit 3 detects the cross point with the low number of lines per 100 lines. However, there are cases where, for example, the reference area is too large for the 150th line and the adjacent cross point is included in the area, so that the cross point cannot be detected suitably. Therefore, the halftone dot feature detection unit 30 of the first modification includes the halftone dot feature detection unit a51 in the 5x5 area and the halftone dot feature detection unit b52 in the 3x3 area, and at least one of the crosspoints is detected. If it is, the cross point is determined.

  Since the halftone dot feature detection unit according to Modification 1 is configured to detect cross-point pixels using the symmetry of the rising edge and the falling edge for each reference area with reference to a plurality of areas having different sizes. From a line number halftone dot to a high line number halftone dot, a halftone dot region can be identified with high accuracy regardless of the number of lines.

(2. Embodiment 2)
(2.1. Halftone dot area identification device according to Embodiment 2)
FIG. 11 is a schematic diagram showing a cross region pattern in which the edge line is oblique. FIG. 12 is an enlarged view of a part of the pattern of FIG.

  In the first embodiment, the vertical edge value and the horizontal edge value are calculated and the halftone dot feature is detected. Therefore, the halftone dot feature is most easily acquired as shown in FIG. 29 arranged at 45 °. It is a halftone dot. Further, since the two-dimensional filter shown in FIG. 5 is used, it can cope with some screen angles and document inclinations. However, as shown in FIGS. 11 and 12, as shown in FIGS. When the pattern is inclined 45 °, the halftone dot feature is most difficult to detect. That is, there are cases where it is not easy to cope with the detection of only the vertical and horizontal edges. Furthermore, in the configuration only for detecting the vertical edge and the horizontal edge, there is a side surface configured so as not to correspond intentionally because the diagonal line is identified and excluded.

  FIG. 13 is a functional block diagram of the halftone dot region identification device according to the second embodiment. The halftone dot region identification device 100 of the second embodiment differs from the first embodiment in that in addition to the horizontal edge detected by the first edge detection unit 101 and the vertical edge detected by the second edge detection unit 102 Thus, a third edge detection 103 and a fourth edge detection 104 for detecting an oblique edge are provided. Moreover, the point provided with the halftone dot feature detection part Y106 which detects a halftone dot feature from the diagonal edge detected by the 3rd and 4th edge detection parts 103 and 104 differs.

  FIG. 14 is a schematic diagram illustrating an example of an oblique edge detection filter used by the third edge detection unit 103 and the fourth edge detection unit 104 illustrated in FIG.

  The third edge detection 103 and the fourth edge detection 104 detect the oblique edge 1 and the oblique edge 2 using two oblique edge detection filters 141 and 142 having different directions shown in FIG.

  The halftone dot feature detection unit X105 detects a cross point of a halftone dot near a screen angle of 45 ° based on the horizontal and vertical edges detected by the first and second edge detection units 101 and 102. The halftone dot feature detection unit Y106 is arranged by rotating the screen angle 45 ° halftone dot by minus 45 ° from the oblique edge 1 and the oblique edge 2 detected by the third and fourth edge detection units 103 and 104. A cross point (halftone dot pattern shown in FIGS. 11 and 12) is detected.

  As for the oblique edge, the cross point is detected in the same manner as that detected in the first embodiment by using the symmetry of the + edge and the − edge in each direction (FIG. 12).

  The adder 107 calculates the logical sum of the cross points detected by the halftone dot feature detection unit X105 and the halftone dot feature detection unit Y106. Density correction 108 and expansion 109 are subjected to density correction processing and expansion processing as in the first embodiment.

(2.2. Effect)
As described above, according to the present embodiment, in addition to the horizontal edge and the vertical edge, the diagonal edge 1 and the diagonal edge 2 orthogonal to the diagonal edge 1 are detected, and similarly, the cross point using the symmetry of the + edge and the − edge. Therefore, for example, a cross point can be detected even at a halftone dot other than a screen angle of 45 ° such as a color halftone dot or a skew.

(2.3. Modification 2)
FIG. 15 is a functional block diagram of a halftone dot region identification device according to a modification of the second embodiment. In the halftone dot region identification device according to the second embodiment, the logical sum of the crossing point of the halftone dot around 45 ° (FIG. 29) and the crossing point of the halftone dot around 90 ° (FIGS. 11 and 12) is obtained before density correction. Implemented. On the other hand, the present modification 2 is different in that a density correction unit is separately provided as shown in FIG. 15 and the addition processing is performed after density correction is performed after detection of the cross point of the angle. That is, as shown in FIG. 15, the density correction unit X207 and the density correction unit Y208 are separately provided, and cross points that are erroneously detected in the character portion are detected as cross points detected from vertical and horizontal edges and cross points detected from diagonal edges. Therefore, it is possible to improve the density correction accuracy with respect to erroneous detection of the character portion.

  Note that when performing dilation processing or cycle determination, the cycle determination may also improve identification accuracy by separately processing the vertical and horizontal edges and the diagonal edges. In such a case, it is desirable to provide each separately like the density correction unit.

(3. Embodiment 3)
(3.1. Overall configuration of halftone dot region identification device)
FIG. 16 is a functional block diagram of the halftone dot region identification device according to the third embodiment. A halftone dot region identification device 300 includes a color conversion unit 301, a C-peak point detection unit 302, a C-cross point detection unit 303, an M-peak point detection unit 305, an M-cross point detection unit 306, and a Y-peak point detection. A unit 308, a Y-cross point detection unit 309, adders 304, 307, 310, 311, a density correction unit 312, and an expansion unit 313.

  The color conversion unit 301 performs color conversion processing on RGB input image data acquired via scanner reading or a network, and converts the input image data into CMY image data. This CMY is a process color CMY. The color conversion process is performed according to the following equation.

C = a0 + a1 x R + a2 x G + a3 x B
M = b0 + b1 × R + b2 × G + b3 × B (Formula 1)
Y = c0 + c1 * R + c2 * G + c3 * B
Here, a0 to a3, b0 to b3, and c0 to c3 are color conversion parameters.

  In order to absorb the nonlinearity of RGB → CMY conversion, it is possible to use hue-division color conversion that switches color conversion parameters for each hue. Further, it is also possible to use LUT color conversion in which CMY values on representative grid points are stored in an LUT (look-up table) and an interpolation operation is performed.

  FIG. 17 is a diagram for explaining the occurrence of rosettes in a 3C gray image. In general color printing originals, some special colors may be used, but most of them use superposition of C (cyan) plate, M (magenta) plate, and Y (yellow) plate. In the third embodiment, the superposition of these colors is decomposed into the original C plate, M plate, and Y plate. The 3C gray image in which the rosette of FIG. 18 is generated is decomposed into each plate.

  Next, for each color of CMY, for example, for cyan, the C peak point detection unit 302 detects a peak, and the C cross point detection unit 303 detects a cross point. Peak point detection and cross point detection are performed for magenta and yellow, respectively.

(3.2. Peak point detector)
FIG. 18 is a schematic diagram for explaining a peak point. 19, 20 and 21 are schematic diagrams illustrating cross points. In the peak point detection, a highlight halftone dot as shown in FIG. 18 or a peak point in a dark halftone dot is detected. However, no peak point is detected at the 50% halftone dot (cross point) in FIG. For the 50% halftone dot in FIG. 19, the cross point detection unit detects that it is a cross point, and detects a halftone dot feature.

  In the peak point detection, the center pixel is detected with reference to a square area 1901 surrounded by a thick solid line in FIG. A region 1802 is an enlarged view of the region 1801 within the solid line in FIG. The peak point 1803 is characterized by a large density difference between the central pixel of the peak point and the peripheral pixels located near the outer frame of the reference area. FIG. 18 illustrates highlight halftone dots, but black and white can be reversed for dark halftone dots.

  On the other hand, when trying to detect a peak point with 50% halftone dots, the dotted line 2001 in FIG. 19 needs to be at least the reference area size. In addition, although the area 1901 in FIG. 20 is enlarged within the dotted line, even when determining the density difference from the peripheral pixel that is the feature of the peak point, there are halftone dots in the entire reference area, Since the density difference is not clear and the halftone dots adjacent to the four corners appear, the density difference in the diagonal direction cannot be determined appropriately, and it is difficult to detect the peak point. Therefore, conventionally, halftone dots in the vicinity of 50% have been covered by peak point detection by detecting peak points with a relaxed and set determination criterion.

  Here, a conventional peak point detection with a relaxed threshold value will be briefly described. For example, the IEICE Transactions Vol. The peak pixel detection method is described in J75-D2 1992-1, “Image area separation method of mixed image of characters / patterns (halftone dots, photographs)”. According to the description, in the peak pixel detection, it is determined from the density relationship with surrounding pixels whether or not the target pixel is an extreme point indicating a peak of density change. In the block consisting of M × M pixels, when “the density level of the central pixel is higher or lower than all other density levels” (condition 1), it is Determine if.

  22 and 23 are diagrams for explaining the density level of the central pixel.

(1) When M = 3 (pixel 2201),
| 2m0−m1−m8 | ≧ ΔmTH and
| 2m0−m2−m7 | ≧ ΔmTH and (Formula 2)
| 2m0−m3−m6 | ≧ ΔmTH and
| 2m0 − m4 − m5 | ≧ ΔmTH

(2) When M = 5 (pixel 2202),
| 2m0−m3−m22 | ≧ ΔmTH and
| 2m0−m8−m17 | ≧ ΔmTH and
・ ・ ・ ・ ・ ・ ・ ・ ・ ・ ・ (Formula 3)
| 2m0−m1−m24 | ≧ ΔmTH and
｜ 2m0 − m7 − m18 | ≧ ΔmTH

  That is, when the absolute value of the density difference between the average value of the two pixel levels and the central pixel located symmetrically with respect to the central pixel is equal to or greater than the threshold value ΔmTH, the central pixel is detected as the peak pixel.

  FIGS. 24-1 and 24-2 are graphs showing an example of the relationship between the pixel position and the density. According to the conventional peak point detection described above, for example, when the conditional expression of the first row of Expression 2 is considered in the peak pixel detection at M = 3, the density difference between the target pixel m0 and the pixels m1 and m8 on both sides is illustrated. Even in the case of the relationship shown in 24-1, m0 is detected as a peak point. That is, it is sufficient that the density difference between m0 and m1 is 1 or more and the above conditional expression is satisfied. In this case, it can be said that almost only the density difference of m8 on one side is effective.

  However, what is originally desired to be detected for a peak point such as the pixel 1801 in FIG. 18 is a pixel having a certain density difference with respect to m1 and m8 on both sides as shown in FIG. 24-2. Therefore, in the peak point detection of this embodiment, instead of the above (Condition 1) in the detection conditions of the conventional peak point detection, “the density level of the central pixel is a predetermined value (see FIG. 24) than all other density levels. (Α) is higher or lower than “(condition 2)”.

  By the way, not only the 50% halftone dot that has been saved by the relaxed reference peak point detection, but also the color halftone dot on which the plates overlap as shown in FIG. . The overlapping of halftone dots destroys the shape of the original isolated dot of the halftone, making it difficult to determine.If peak point detection is performed to see the density difference on both sides as in this embodiment, color mixing will occur. There was a demerit that the detection accuracy at the color halftone dot deteriorated. However, in the present embodiment, the shape of the isolated dot can be substantially restored by converting it to process color CMY before the peak point detection, so that the peak point detection accuracy in the mixed color halftone dot can be ensured.

(3.3. Cross point detection unit, density correction unit, and expansion unit)
The cross point detection unit refers to a square area 1902 surrounded by a thick solid line in FIG. 19 and detects the center pixel. As shown in FIG. 19, the cross point detection has an advantage that it can be detected in a reference region smaller than the peak point in addition to the effect of suppressing erroneous detection in the character portion. FIG. 21 is an enlarged view of the solid line 1902.

  Cross point detection is performed, for example, by the cross point detection method according to the first embodiment described above. That is, the cross point is detected as a halftone dot feature by the symmetry of the rising edge and falling edge of the horizontal edge and the symmetry of the rising edge and falling edge of the vertical edge. The cross point detection method is not limited to this, and can also be realized by performing detection based on the maximum value and the minimum value of the density described in Patent Document 1 mentioned above.

  As for the cross point detection, as in the case of the peak point detection, the detection accuracy in the mixed color halftone dot is improved by detecting in the CMY version than in the case of detecting by RGB.

  Adders 304, 307, and 310 take the logical sum of the detected peak points and cross points, respectively.

  The adder 311 takes the logical sum of each version of the logical sum of the peak point and the cross point detected for each CMY version.

  The density correction unit 312 detects an excessively detected peak point or cross point that cannot be a halftone dot region, such as a cross point that exists in isolation, with respect to the logical sum of the peak point and cross point generated by the adder 311. Is eliminated depending on the density. Here, not only the density but also the period of peak points and / or cross points may be determined to determine whether or not to eliminate them.

  The expansion unit 313 identifies the halftone dot region by including the halftone dot region and the halftone dot region by expanding the peak point and / or the cross point after density correction.

(3.4. Halftone dot region identification procedure)
FIG. 25 is a flowchart illustrating a halftone dot region identification procedure according to the third embodiment. The function of each unit described above is executed by taking the following procedure as a halftone dot region identification procedure. The color conversion unit 301 converts input RGB into CMY (step S201). Each peak point detection unit 302, 305, 308 detects a peak point for each color (step S202). Each cross point detection part 303,306,309 detects the cross point for every color (step S203). The adders 304, 307, and 310 perform a logical sum of the detected peak points and cross points for each CMY (step S204). The adder 311 further takes all the logical sums for the logical sums taken for each CMY (step S205). The density correction unit 312 performs correction processing on the cross points and peak points detected according to the density (step S206). The expansion unit 313 performs expansion processing on the cross points and / or peak points subjected to correction processing by the density correction unit 312 and identifies them as halftone dot regions (step S207).

(3.5. Effect)
As described above, according to the present embodiment, after the color conversion from the input color RGB data to the process color CMY, the peak point and the cross point are detected for each plate. Similar to the halftone dots, the peak point detection and the cross point detection can identify a halftone dot region with high accuracy while suppressing erroneous detection particularly in a character portion, for example.

(3.6. Modification 3)
FIG. 26 is a functional block diagram of a modification of the third embodiment. In the third embodiment, density correction and expansion processing are performed after the peak points and cross points detected for each CMY plate are added, but in this third modification, the peak points and cross points of each color of CMY are used. This is different in that density correction is performed for each color before adding all. That is, the difference is that each has a C density correction unit 431, an M density correction unit 432, and a Y density correction unit 433.

  With this configuration, the density correction parameters can be set differently for each plate. For example, a difference in color conversion accuracy for each CMY plate can be absorbed with higher accuracy. Of course, the peak point detection and cross point detection parameters can also be absorbed, but if both are absorbed, the effect is enhanced. As a result, it has a favorable effect on the improvement of the density correction accuracy for erroneous detection of the character part.

  For example, when the C peak point is present in isolation and the M peak point is also present nearby, the density is lower when viewed from each plate, and it is easier to eliminate the isolated peak.

  In the case of performing expansion processing or cycle determination, it may be possible to improve the identification accuracy if the cycle determination is performed separately for each plate. Therefore, a configuration in which the cycle determination is performed separately for each plate is also possible.

  In this embodiment, the density correction is performed after taking the OR of the peak point and the cross point for each color. However, in order to perform finer control, the peak point and the cross point are determined before taking the OR. A configuration in which density correction is performed separately is possible.

(3.7. Modification 4)
In the third embodiment and the third modification, the highlight halftone dot, the dark halftone dot, and the intermediate density (near 50%) halftone dot are changed to the peak point and It was configured to switch appropriately whether to cover at either of the cross points. Referring to FIG. 16, the adder 304 takes the OR of the results of the C peak point detection 302 and the C cross point 303.

  FIG. 27 is a diagram for explaining a main part of a halftone dot region identification device according to another modification of the third embodiment. The halftone dot region identification device according to the modification 4 includes, for C, for example, an average density calculation unit 541 and a selector 544 in addition to the C-peak point detection unit 542 and the C-cross point detection unit 543.

  In the halftone dot region identification device according to the fourth modification, as shown in FIG. 27, the average density calculation unit 541 calculates the average density in the predetermined reference area including the target pixel, and the calculated average density value is selected by the selector 544. The selector 544 selects either the peak point or the cross point according to the average density. Threshold processing is performed on the average density. If the density is low or high, a peak point is selected, and if it is an intermediate density, a cross point is selected. With this configuration, switching of the assigned density halftone dot between the peak point and the cross point can be performed more easily than the parameter adjustment for peak point detection and cross point detection, and the operation becomes simpler.

  In addition, for example, a configuration in which a low / medium concentration near the boundary between a low concentration and an intermediate concentration is further determined and the peak point and the cross point are ORed in the low / medium concentration region is also possible as another modification.

  In addition, 1. 1. Configuration using another halftone dot feature detection suitable for detection of highlight halftone dots and dark halftone dots instead of peak point detection; There is also a realization means by a configuration using another halftone dot feature detection suitable for the detection of the intermediate density halftone dot instead of the cross point detection. For example, a configuration using a method of detecting halftone dots by pattern matching described in JP-A-63-279665 instead of peak point detection, and JP-A-4-14378 instead of cross-point detection. (White background detection can detect halftone dots other than highlight halftone dots as non-white backgrounds).

(4. Hardware configuration, recording medium)
FIG. 31 is a diagram illustrating a hardware configuration example of the halftone dot region identification device according to the embodiment. The above-described halftone dot region identification device can be realized by executing a program prepared in advance on a computer system such as a personal computer or a workstation. The entire computer 700 is controlled by a CPU (Central Processing Unit) 701. A ROM (Read Only Memory) 702, a RAM (Random Access Memory) 703, a hard disk drive (HDD: Hard Disk Drive) 704, a graphic processing device 705, and an input interface 706 are connected to the CPU 701 via a bus 707. The ROM 702 and the RAM 703 store at least a part of an OS (Operating System) program and application programs to be executed by the CPU 701. The RAM 703 stores various data necessary for processing by the CPU 701. The HDD 704 stores an OS, various driver programs, application programs, detected data, and the like.

  A monitor 711 is connected to the graphic processing device 705. The graphic processing device 705 displays an image on the screen of the monitor 711 in accordance with a command from the CPU 701. A keyboard 712 and a mouse 713 are connected to the input interface 706. The input interface 706 transmits a signal sent from the keyboard 712 and the mouse 713 to the CPU 701 via the bus 707.

  With the hardware configuration as described above, the processing functions of the present embodiment can be realized. In order to implement the present embodiment on the computer 700, a driver program is mounted on the computer 700.

  The halftone dot area identification program executed by the halftone dot area identification device according to the present embodiment is a file in an installable or executable format and is read by a computer such as a CD-ROM, floppy (R) disk, or DVD. Provided by being stored in a possible recording medium.

  The halftone dot area identification program of the present embodiment may be configured to be provided and distributed by storing it on a computer connected to a network such as the Internet and downloading it via the network.

  As described above, the halftone dot area identification device and the halftone dot area identification method according to the present invention are useful for an image processing apparatus and are particularly suitable for a copying machine.

3 is a functional block diagram of the halftone dot region detection apparatus according to Embodiment 1. FIG. It is a functional block diagram of the 1st edge detection part and 2nd edge detection part which were shown in FIG. 5 is a flowchart illustrating a halftone dot region detection operation according to the first embodiment. It is a schematic diagram explaining the edge part in an image. It is a schematic diagram which shows an example of the filter for edge detection. FIG. 2 is a functional block diagram of a halftone dot feature detection unit shown in FIG. 1. It is a schematic diagram which shows an example of the mask used with the halftone dot characteristic detection part of FIG. It is a figure explaining the pattern of a cross point. It is a schematic diagram explaining a peak point. 12 is a functional block diagram of a halftone dot feature detection unit according to Modification 1. FIG. It is a schematic diagram which shows the cross area | region pattern whose edge line in an image is diagonal. It is the figure which expanded a part of pattern of FIG. FIG. 6 is a functional block diagram of a halftone dot area identification device according to a second embodiment. FIG. 14 is a schematic diagram illustrating an example of an oblique edge detection filter used by the third edge detection unit and the fourth edge detection unit illustrated in FIG. 13. FIG. 10 is a functional block diagram of a halftone dot region identification device according to a modification of the second embodiment. FIG. 10 is a functional block diagram of a halftone dot region identification device according to a third embodiment. It is a figure explaining generation | occurrence | production of the rosette in 3C gray image. It is a schematic diagram explaining the peak point in an image. It is a schematic diagram explaining the cross point in an image. It is a schematic diagram explaining the cross point in an image. It is a schematic diagram explaining the cross point in an image. It is a figure explaining the density level of a center pixel. It is a figure explaining the density level of a center pixel. It is a graph which shows an example of the relationship between a pixel position and a density | concentration. It is a graph which shows an example of the relationship between a pixel position and a density | concentration. 10 is a flowchart illustrating a halftone dot region identification procedure according to the third embodiment. FIG. 10 is a functional block diagram of a modified example of the third embodiment. It is a figure explaining the principal part of the halftone-dot area | region identification apparatus by the other modification of Embodiment 3. FIG. It is a figure explaining the halftone image by a peak point. It is a figure explaining the halftone image by a cross point. It is a figure explaining the halftone image by a cross point. It is a figure which shows the hardware structural example of the halftone-dot area | region identification apparatus by this Embodiment.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 1st edge detection part 2 2nd edge detection part 3, 51, 52 Halftone feature detection part 4, 108 Density correction part 5, 109, 210 Expansion part 11 Horizontal edge detection filter 12, 22 Code | symbol determination part 13 , 23 Multiplier 14, 24 Selector 15, 17 Absolute value conversion unit 16 Subtractor 18 Comparator 21 Vertical edge detection filter 31-34, Mask 35, 36, 39 Sign inversion unit 37, 40 Minimum value selection unit 38, 41 Threshold determination unit 42, 53 Adder 101, 201 First edge detection unit 102, 202 Second edge detection unit 103, 203 Third edge detection unit 104, 204 Fourth edge detection unit 105, 205 network Point feature detection unit X
106, 206 Halftone dot feature detection unit Y
107, 209 Adder 207 Density correction unit X
208 Density correction part Y
301, 421 Color converter 302, 422, 542 C-peak point detector 303, 423, 543 C-cross point detector 305, 425 M-peak point detector 306, 426 M-cross point detector 308, 428 Y -Peak point detection unit 309, 429 Y-cross point detection unit 304, 307, 310, 311, 424, 427, 430, 434 Adder 312, 431-433 Density correction unit 313, 435 Expansion unit 541 Average density calculation unit 544 Selector 700 computer

Claims (26)

First edge detection means for detecting a rising edge and a falling edge in a first direction in the image data with reference to a region in the image data;
A second edge detecting means for detecting a rising edge and a falling edge in a second direction orthogonal to the first direction in the image data with reference to the region;
The edges detected by the first and second edge detecting means are referred to in a predetermined area, the sum of rising edges in the first direction in the referred predetermined area, Calculate the sum of the falling edges, the sum of the rising edges in the second direction, and the sum of the falling edges in the second direction, and unify the sign of each calculated value, A halftone dot feature detecting means for detecting whether or not a minimum value of the threshold value is equal to or greater than a threshold value for judging edge bias and detecting a pixel having a halftone dot feature based on the determination result ;
A halftone dot region identification device comprising:   2. A halftone dot according to claim 1, wherein the halftone dot feature detection means detects a cross dot pixel that is a central pixel between nearest halftone dot dots as the pixel having the halftone dot feature. Area identification device.   The halftone dot region identification apparatus according to claim 1 or 2, wherein each of the first edge detection means and the second edge detection means detects an edge using a first-order differential filter. The halftone dot feature detection means includes:
When the edge data detected by the first edge detection means and the edge data detected by the second edge detection means are compared, and the absolute values of the indices representing the sizes of both edges are determined to be substantially equal The halftone dot region identification device according to claim 1, wherein the halftone dot region identification device is determined not to be a cross point pixel. The halftone dot feature detection means refers to a plurality of predetermined reference areas having different sizes , and sums rising edges in the first direction and falling in the first direction for each of the plurality of predetermined reference areas. The sum of edges, the sum of rising edges in the second direction, and the sum of falling edges in the second direction are calculated, the signs of the calculated values are unified, and the smallest of the values obtained by unifying the signs It is determined whether a value is more than the threshold value, and based on the determination result , a pixel having a halftone dot feature is detected. Halftone dot area identification device. further,
Third edge detecting means for detecting a rising edge and a falling edge in a third direction which is an intermediate direction between the first direction and the second direction;
A fourth edge detecting means for detecting a rising edge and a falling edge in a fourth direction orthogonal to the third direction,
The halftone dot feature detection means includes:
The edges detected by the third and fourth edge detecting means are referred to in the predetermined area, and the sum of rising edges in the third direction in the referred predetermined area, the third direction The sum of the falling edges, the sum of the rising edges in the fourth direction, and the sum of the falling edges in the fourth direction are calculated, and the signs of the calculated values are unified. It is determined whether or not the minimum value is equal to or greater than the threshold, and the determination result using the edges in the first direction and the second direction, the edges in the third direction and the fourth direction are determined. 6. The halftone dot region identification device according to claim 1, wherein a pixel having a halftone dot feature is detected based on one of the determination results used . The halftone dot feature detection means includes:
First halftone dot feature detection means for detecting a pixel having a halftone dot feature based on a determination result using edges in the first direction and the second direction ;
Second halftone dot feature detection means for detecting pixels having a halftone dot feature based on the determination result using the edges in the third direction and the fourth direction ,
further,
First correction means for correcting halftone dot feature pixel data detected by the first halftone dot feature detection means;
Second correction means for correcting the halftone dot feature pixel data detected by the second halftone dot feature detection means;
The halftone dot region identification device according to claim 6, comprising: A first edge detecting step of detecting a rising edge and a falling edge in a first direction in the image data with reference to a region in the image data;
A second edge detecting step of detecting a rising edge and a falling edge in a second direction orthogonal to the first direction in the image data with reference to the region;
The edge data detected by the first edge detection step and the second edge detection step is referred to in a predetermined region, and the sum of rising edges in the first direction in the predetermined region referred to, Calculate the sum of the falling edges in the first direction, the sum of the rising edges in the second direction, and the sum of the falling edges in the second direction, unify the sign of each calculated value, and determine the constant step minimum value of the unified value you determine whether a threshold value or more to determine the deviation of the edge,
A halftone feature detection step of detecting a pixel having a dot feature based on the determination result of the previous SL-size constant step,
A halftone dot region identification method comprising:   9. The halftone dot according to claim 8, wherein the halftone dot feature detecting step detects a cross dot pixel that is a central pixel between nearest halftone dot dots as the pixel having the halftone dot feature. Area identification method.   10. The halftone dot region identification method according to claim 8, wherein the first edge detection step and the second edge detection step each detect an edge using a first-order differential filter. The dot feature detection step includes
When the edge data detected by the first edge detection step and the edge data detected by the second edge detection step are compared, and it is determined that the absolute values of the indices indicating the sizes of both edges are substantially equal The halftone dot region identification method according to any one of claims 8 to 10, wherein it is determined that the pixel is not a cross point pixel. Before SL-size constant step, with reference to the plurality of predetermined reference region having different sizes, the sum of the first direction of a rising edge for each of the plurality of predetermined reference region, the falling edge of the first direction , The sum of the rising edges in the second direction, and the sum of the falling edges in the second direction, and unifying the sign of each calculated value, and the minimum value among the unified values Whether or not is greater than or equal to the threshold,
12. The halftone dot feature detecting step detects pixels having a halftone dot feature based on a determination result determined for each of the plurality of predetermined reference regions. The halftone dot region identification method according to one. further,
A third edge detecting step for detecting a rising edge and a falling edge in a third direction which is an intermediate direction between the first direction and the second direction;
A fourth edge detecting step for detecting a rising edge and a falling edge in a fourth direction orthogonal to the third direction,
Before SL-size constant process,
The edge data detected by the first edge detection step and the second edge detection step is referred to in a predetermined region, and the sum of rising edges in the first direction in the predetermined region referred to, Calculate the sum of the falling edges in the first direction, the sum of the rising edges in the second direction, and the sum of the falling edges in the second direction, unify the sign of each calculated value, A first determination step of determining whether or not a minimum value among the unified values is equal to or greater than the threshold ;
The edge data detected by the third edge detection step and the fourth edge detection step is referred to in the predetermined region, and the sum of rising edges in the third direction in the predetermined region referred to, Calculating the sum of the falling edges in the third direction, the sum of the rising edges in the fourth direction, and the sum of the falling edges in the fourth direction, unifying the sign of each calculated value, And a second determination step of determining whether or not a minimum value among the unified values is equal to or greater than the threshold value ,
The halftone dot feature detection step detects pixels having a halftone dot feature based on one of the determination result of the first determination step and the determination result of the second determination step. The halftone dot region identification method according to any one of claims 8 to 12. The dot feature detection step includes
A first halftone dot feature detection step for detecting pixels having a halftone dot feature based on a determination result of the first determination step;
A second halftone dot feature detection step for detecting pixels having a halftone dot feature based on the determination result of the second determination step;
A first correction step of correcting the halftone dot feature pixel data detected by the first halftone dot feature detection step;
A second correction step of correcting the halftone dot feature pixel data detected by the second halftone dot feature detection step;
The halftone dot region identification method according to claim 13, comprising: further,
Color conversion means for converting input image data comprising a plurality of color signals into image data comprising a plurality of predetermined color signals having attributes different from those of the plurality of color signals;
Peak point detection means for detecting a peak point that is a central pixel of halftone dots from the image data after color conversion processing by the color conversion means;
Halftone dot area identifying means for identifying a halftone dot area in the input image data,
The first edge detection means detects a rising edge and a falling edge in a first direction in the image data with reference to a region in the image data after the color conversion processing by the color conversion means;
The second edge detection means refers to the region in the image data after color conversion processing, detects a rising edge and a falling edge in the second direction in the image data,
The halftone dot feature detection means is a central pixel between nearest halftone dots as a pixel having the halftone dot feature based on a determination result using the edges in the first direction and the second direction. Detect cross points,
The halftone dot region identification device according to claim 1, wherein the halftone dot region identification unit identifies a halftone dot region in the input image data based on the peak point and the cross point .   The color conversion unit converts input image data including a plurality of color signals into image data including a plurality of predetermined color signals having attributes different from those of the plurality of color signals including signals corresponding to process colors. The halftone dot region identification device according to claim 15. The peak point detection means detects at least low density and high density halftone dots,
The halftone dot region identification device according to claim 15 or 16, wherein the halftone dot feature detection means detects at least a halftone dot having an intermediate density. The peak point detection means detects the peak point for each of the predetermined plurality of color signals converted by the color conversion means,
18. The halftone dot feature detection unit detects the cross point for each of the predetermined plurality of color signals converted by the color conversion unit. A halftone dot region identification device according to claim 1. Furthermore, a correction unit that performs correction for each of the plurality of converted colors based on one or more of the detected data of the peak point and the cross point ,
The halftone dot region identifying unit is for identifying a halftone dot region based on the data corrected by the correcting unit. Area identification device.   The peak point detection means determines whether or not the density difference between the target pixel and surrounding pixels surrounding the target pixel is equal to or greater than a predetermined threshold value. The halftone dot region identification device according to any one of claims 15 to 19, wherein it is determined that it is not. further,
A color conversion step of converting input image data composed of a plurality of color signals into image data composed of a plurality of predetermined color signals having attributes different from those of the plurality of color signals;
From the image data after the color conversion process in the color conversion step, a peak point detection step for detecting a peak point that is a central pixel of halftone dots,
Anda dot region identification step of identifying the halftone dot area in the entering force image data,
The first edge detection step refers to a region in the image data after the color conversion processing by the color conversion step, detects a rising edge and a falling edge in the first direction in the image data,
The second edge detection step refers to the region in the image data after color conversion processing, detects a rising edge and a falling edge in the second direction in the image data,
The halftone dot feature detection step detects a cross point that is a central pixel between nearest halftone dot dots as a pixel having a halftone dot feature based on the determination result of the determination step,
9. The halftone dot region identifying method according to claim 8, wherein the halftone dot region identifying step identifies a halftone dot region in the input image data based on the peak point and the cross point .   The color conversion step converts input image data including a plurality of color signals into image data including a plurality of predetermined color signals having attributes different from those of the plurality of color signals including signals corresponding to process colors. The halftone dot region identification method according to claim 21, wherein: The peak point detection step detects at least low density and high density halftone dots,
The halftone dot region identifying method according to claim 21 or 22, wherein the halftone dot feature detecting step detects at least a halftone dot having an intermediate density. The peak point detection step detects the peak point for each of the predetermined plurality of color signals converted by the color conversion step.
24. The halftone dot feature detecting step detects the cross point for each of the predetermined plurality of color signals converted by the color converting step. 2. A halftone dot region identifying method described in 1. And a correction step of performing correction for each of the plurality of colors converted in the color conversion step based on one or more of the detected data of the peak point and the cross point ,
25. The halftone dot region identifying step is a step of identifying a halftone dot region based on the data that has been subjected to correction processing in the correction step. Halftone dot region identification method.   The peak point detection step determines whether or not the density difference between the target pixel and surrounding pixels surrounding the target pixel is equal to or greater than a predetermined threshold value. The halftone dot region identification method according to any one of claims 21 to 25, wherein the halftone dot region identification method according to any one of claims 21 to 25 is determined.

Images