JP6641982B2 - Image processing apparatus, image processing method, and program - Google Patents

Description

  The present invention relates to an image processing device, an image processing method, and a program.

  As an image compression technique for achieving both high compression and high image quality, a technique called high compression PDF (Portable Document Format) is known. High-compression PDF is a technology that separates an image into a line image such as a character and other areas and performs appropriate compression processing on each of them to achieve both high compression and high image quality.

  As a method of detecting a line drawing from an image, for example, focusing on the fact that a character that is a representative line drawing often forms a line, extracting a connected component as a character candidate, and then extracting a color of the connected component A method is known in which a character line is extracted by using character strings and positional relationships, and connected components included in the character line are detected as characters (for example, see Patent Document 1). In addition, various methods are known, such as a method of detecting an edge constituting a line image from an image by pattern matching of connected pixels (for example, see Patent Document 2).

  However, line drawings detected from an image by a conventional method may include a line drawing which, when subjected to a compression process for a line drawing, may rather lower the compression ratio. Further, in some cases, a portion which is not detected as a line drawing in the conventional method includes a region where it is desirable to perform compression processing for the line drawing. For this reason, if the line drawing detected by the conventional method is uniformly subjected to line drawing compression processing, efficient compression may not always be performed.

In order to solve the above-described problem, the present invention provides a detecting unit for detecting a candidate of a target area to be subjected to a first compression process from an image, and for each of the detected candidates, the number of colors possessed by the candidate. Calculating means for calculating the number of colors representing the image data; determining means for determining whether or not the candidate is the target area based on at least the number of colors of the candidate; and the first compression for the target area. Compression processing means for performing processing and performing second compression processing different from the first compression processing on an area other than the target area included in the image , wherein the calculation means For each of the candidates, a background color that is a color of an area adjacent to the candidate and a color of the candidate are further calculated, and the determining unit determines the number of colors of the candidate and the background color of the candidate. , Based on the color of the candidate, To determine whether or not the serial target area.

  ADVANTAGE OF THE INVENTION According to this invention, there exists an effect that the compression efficiency with respect to an image can be improved.

FIG. 1 is a schematic diagram illustrating a processing procedure of the high-compression PDF. FIG. 2 is a block diagram illustrating an example of a hardware configuration of the image processing apparatus according to the embodiment. FIG. 3 is a block diagram illustrating a functional configuration example of the image processing apparatus according to the embodiment. FIG. 4 is a block diagram illustrating a specific example of the detection unit. FIG. 5 is a flowchart illustrating the flow of the operation of the image processing apparatus according to the embodiment. FIG. 6 is a flowchart illustrating an example of a process performed by the first detection unit. FIG. 7 is a flowchart illustrating an example of a process performed by the second detection unit. FIG. 8 is a flowchart illustrating an example of a process performed by the calculation unit. FIG. 9 is a flowchart illustrating an example of a process performed by the determining unit. FIG. 10 is a flowchart illustrating an example of a process performed by the compression processing unit.

  Hereinafter, specific embodiments of an image processing apparatus, an image processing method, and a program according to the present invention will be described in detail with reference to the accompanying drawings. The embodiment described below is an example of application to an image processing apparatus that generates a high-compression PDF file. However, the present invention is not limited to this example, and can be effectively applied to various image processing apparatuses that require efficient image compression.

<High compression PDF>
First, prior to a specific description of the present embodiment, an outline of the high compression PDF will be described. High compression PDF is an image compression technique for generating a high compression PDF file from an image including a line drawing such as a character. Here, the line drawing refers to a character and an object represented by a line that is desirably handled in the same manner as the character. Even if it is an object represented by a line, an object that is desirably handled as a picture is included in the picture instead of the line drawing. A picture is an object other than a line drawing, that is, an object represented by a halftone dot such as a photograph, or an object such as a figure that is not desirably handled like a character.

  FIG. 1 is a schematic diagram illustrating a processing procedure of the high-compression PDF. In order to generate a high-compression PDF file, first, a first image layer Im1 which is a binary image consisting only of a line image and a color of the line image are obtained from an image to be processed (hereinafter referred to as an “input image”) Im0. Is generated, and a third image layer Im3 which is a multi-valued image expressing a pattern other than a line drawing and a background is generated. Then, the first image layer Im1 and the second image layer Im2 are subjected to compression processing suitable for compressing a line image, and the third image layer Im3 is suitable for compression of a pattern or background. Perform compression processing. Thereafter, the first image layer Im1 subjected to the compression processing, the second image layer Im2 subjected to the compression processing, and the third image layer Im3 subjected to the compression processing are converted into, for example, one PDF format. By integrating them on one image file, a high-compression PDF file FIm corresponding to the input image Im0 is generated.

  The compression process performed on the first image layer Im1 is, for example, a compression process on a binary image using an encoding method such as MMR. The compression processing performed on the second image layer Im2 is, for example, compression processing on a multi-valued image using a coding method such as JPEG with a lower resolution than the compression processing on the third image layer Im3. Since the compression process for the first image layer Im1 and the compression process for the second image layer Im2 are common in that they are compression processes suitable for compressing line drawings, these processes will be collectively referred to as “first This is called “compression processing”. On the other hand, the compression process performed on the third image layer Im3 is, for example, a compression process using a coding method such as JPEG for a multi-valued image that has a higher resolution than the compression process on the second image layer Im2. is there. The compression process for the third image layer Im3 is a compression process suitable for compressing a picture or a background. Therefore, hereinafter, the “second compression process” will be distinguished from the first compression process suitable for the compression of a line drawing. Call. Note that the above-described encoding method is an example, and a compression process using an encoding method different from the above-described example may be performed.

  In the high-compression PDF, as described above, the input image Im0 to be processed is separated into a line drawing area and other picture and background areas, and the first compression processing is performed on the line drawing area. By applying the second compression process to a pattern or background area other than the line drawing, the compression efficiency is increased. Here, the compression efficiency indicates how much the compression ratio could be increased without impairing the image quality (reproducibility) when the image was reproduced, and a high compression ratio was obtained while maintaining the reproducibility. If so, efficient compression has been performed.

  The above-mentioned high compression PDF can be variously modified. For example, the above-described first image layer Im1 may be divided into an image layer composed of only black line drawings and an image layer composed of only chromatic or white line drawings. Further, a configuration may be adopted in which the color of the line drawing is provided as other information corresponding to the coordinates of the line drawing without having the second image layer Im2 expressing the color of the line drawing.

<Overview of Embodiment>
In order to appropriately generate the high-compression PDF file FIm corresponding to the input image Im0, a line image to be subjected to the first compression processing is accurately detected from the input image Im0, and the input image Im0 is extracted from the line image region and other patterns. And must be properly separated from the background area. However, in the existing method of detecting a line image from the input image Im0, there are cases where objects detected as a line image include those that, if subjected to the first compression processing, will rather decrease the compression efficiency. Further, if such an object is adjusted so as not to be detected as a line drawing, an object for which the first compression processing is desired may not be detected as a line drawing.

  Therefore, in the present embodiment, an object detected as a line drawing by the existing method or an improvement of the existing method is treated as a line drawing candidate to be subjected to the first compression processing. That is, in the present embodiment, candidates for a line drawing (one mode of the “target area”) (hereinafter, referred to as “line drawing candidates”) are detected from the input image Im0. Then, by comparing the detected line drawing candidate with a predetermined condition, it is determined whether or not the line drawing is subjected to the first compression processing. That is, of the detected line image candidates, only line image candidates satisfying a predetermined condition are determined as line images to be subjected to the first compression processing.

  Here, the matching of the predetermined condition uses at least the number of colors of the line drawing candidate calculated for each detected line drawing candidate. Here, the number of colors is an index indicating the number of colors of the line drawing candidate. For example, the line drawing candidate is represented by a single color (the number of colors is 1) or is represented by a plurality of colors (colors). It is used to specify whether the number is 2 or more. For example, a portion where a red character or a blue character overlaps a black character like a written office document or a corrected answer sheet, or a portion where a seal is stamped on a gray “mark” character, etc. When two line drawing candidates are detected, the number of colors of those line drawing candidates is two.

  In addition, the matching of the predetermined condition may use the background color of the line drawing candidate in addition to the number of colors of the line drawing candidate, or may use the color of the line drawing candidate in addition to the number of colors and the background color of the line drawing candidate. Is also good. The background color of a line drawing candidate is the color of an area adjacent to the line drawing candidate. Further, in the collation of the predetermined condition, the aspect ratio of the circumscribed rectangle of the line drawing candidate, the line width of the line drawing candidate, and the like may be used in combination with the information on the color.

  In the present embodiment, as described above, a line image candidate that satisfies a predetermined condition is determined as a line image to be subjected to the first compression process from among the line image candidates detected from the input image Im0. Then, the first image layer Im1 and the second image layer Im2 expressing the line image determined in this way are subjected to the first compression processing, and the third image expressing the other patterns and backgrounds. The second compression process is performed on the layer Im3 to generate a high compression PDF file FIm. Therefore, according to the present embodiment, it is possible to improve the efficiency of compression of the input image Im0 and generate a high-quality high-compression PDF file FIm.

<Calculation of color>
In the present embodiment, as described above, in order to determine whether or not the line drawing candidate detected from the input image Im0 is a line drawing to be subjected to the first compression processing, the number of colors of the line drawing candidate, the background color of the line drawing candidate , And color information such as the color of a line drawing candidate. Such color calculation can be realized by various methods. In the present embodiment, for example, it is assumed that HSV data converted from RGB data of the input image Im0 is used. HSV is a color space that expresses colors by hue (H), saturation (S), and lightness (V).

  Here, a specific example of the color calculation method according to the present embodiment will be described. Here, it is assumed that the RGB data of the input image Im0 is represented by 8 bits for each color of RGB (value range: 0 to 255). The value range of S and V of the HSV data is assumed to be 0 to 255 similarly to the RGB data. Since H in the HSV data represents a hue, it takes a value in the range of 2π (value range: 0 to 359).

Assuming that the maximum values of R, G, and B of the RGB data are MAX and the minimum values are MIN, the RGB data can be converted to HSV data using the following equations (1) to (5).
When MAX = R: H = 60 × ((GB) ÷ (MAX-MIN)) (1)
When MAX = G: H = 60 × ((BR) ÷ (MAX-MIN)) + 120 (2)
When MAX = B: H = 60 × ((R−G) ÷ (MAX−MIN)) + 240 (3)
However, when H <0, 360 is added. When R = G = B, the value of H is undefined.
S = ((MAX−MIN) ÷ MAX) × 255 (4)
However, when R = G = B, S = 0.
V = MAX (5)

  In the present embodiment, for example, for chromatic colors, there are six types of red, yellow, green, cyan, blue, and magenta according to the value of hue (H), and 11 types according to the value of lightness (V). The colors are classified into 66 types that are combined with the (equal interval) classification. Achromatic colors are classified into 11 types (black, white, and equally spaced between them) according to the value of lightness (V). Note that this classification is merely an example, and the number of color classifications and the like can be arbitrarily set. Further, the chromatic colors may be classified only by the value of the hue (H) without using the value of the lightness (V). Whether it is a chromatic color or an achromatic color can be determined using, for example, a value of saturation (S). For example, if S ≦ 5, it is determined to be achromatic, and if S> 5, it is determined to be chromatic. Further, it is determined whether the input image Im0 is a color image or a monochrome image by using the existing technology called ACS (Auto Color Selection), and if the input image Im0 is a monochrome image, all colors are determined to be achromatic. You may.

  The number of colors of the line drawing candidate can be calculated, for example, as follows. First, the RGB data of each pixel constituting the line drawing candidate is converted into HSV data, and the average value of the HSV data is calculated. Then, it is checked whether or not a pixel (hereinafter, referred to as a “discrete pixel”) whose difference from the average value (for example, the Euclidean distance in the HSV color space) is equal to or more than a predetermined value exists in the line drawing candidate. If there is, the ratio of the number of discrete pixels to the total number of line drawing candidates is further calculated. If there are no discrete pixels in the line drawing candidate, or if the ratio of the number of discrete pixels to the total number of line drawing candidates is less than a predetermined value, the number of colors of the line drawing candidate is set to one. On the other hand, when there are discrete pixels in the line drawing candidate and the ratio of the number of discrete pixels to the total number of pixels of the line drawing candidate is equal to or more than a predetermined value, it is estimated that the number of colors of the line drawing candidate is two or more. If the number of colors of the line drawing candidate is estimated to be 2 or more, for example, a histogram representing the distribution of the discrete pixels in the HSV color space is calculated, and the number of colors of the line drawing candidate is calculated based on the number of peaks appearing in the histogram. I do.

  If the number of colors of the line drawing candidate is 1, the average value of the HSV data of each pixel constituting the line drawing candidate is calculated as the color of the line drawing candidate. If the number of line drawing candidates is two or more, the average value of the HSV data is calculated for each color. The colors of the line drawing candidates calculated in this way can be classified into any of the above 66 types if they are chromatic colors, and can be classified into any of the above 11 types if they are achromatic colors. For example, regarding the value of the hue (H) when the color of the line drawing candidate is a chromatic color, if 0 ≦ H <30 or 330 ≦ H <360, it is red, if 30 ≦ H <90, yellow, and 90 ≦ H <. The color of the line drawing candidate can be classified such as 150 for green, 150 for H <210 for cyan, 210 for H <270 for blue, and 270 for H <330 for magenta.

  As the background color of the line drawing candidate, an area of a predetermined size and shape adjacent to the line drawing candidate (there may be a gap of about one pixel) is selected, and the average value of the HSV data of each pixel in this area is calculated. The background color of the line drawing candidate. As with the line drawing candidate colors, the background colors of the line drawing candidates can be classified into any of the above 66 types if they are chromatic colors, and can be classified into any of the above 11 types if they are achromatic colors. If the input image Im0 is an image obtained by reading a document with a scanner and the image processing apparatus has a function of detecting the background color of the document (the original color of the document), the detected background color of the document is used as a line drawing candidate. The background color may be used.

  The example described above is an example of a method of calculating the number of colors of the line drawing candidate, the background color of the line drawing candidate, the color of the line drawing candidate, and the like. The color of a line drawing candidate may be calculated. For example, the RGB data of the input image Im0 is mapped to the Munsell color system by table conversion or the like, and the number of color of the line image candidate, the background color of the line image candidate, the color of the line image candidate, etc. using the hue and lightness in the Munsell color system. May be calculated.

<Configuration of image processing device>
Next, the image processing apparatus according to the present embodiment will be specifically described. FIG. 2 is a block diagram illustrating a hardware configuration example of the image processing apparatus 1 according to the present embodiment. The image processing apparatus 1 can use, for example, a computer system such as a PC (personal computer) as hardware. That is, for example, as shown in FIG. 2, the image processing apparatus 1 includes a processor such as a CPU 101, storage devices such as a RAM 102, a ROM 103, and an HDD 104, and a network I / F 105 that is a communication interface connected to a network such as a LAN. And these are connected via a bus 110.

  The image processing apparatus 1 of the present embodiment acquires an image to be processed (input image Im0) from a scanner or a host computer connected to a network via the network I / F 105, for example. Then, the image processing apparatus 1 generates the high-compression PDF file FIm by processing the input image Im0, stores the generated high-compression PDF file FIm in the HDD 104, or connects to the network via the network I / F 105. Or send it to the designated host computer. The function of generating the high-compression PDF file FIm from the input image Im0 is realized, for example, by the CPU 101 executing a predetermined program stored in the ROM 103, the HDD 104, or the like, using the RAM 102 as a work area.

  Note that the image processing apparatus 1 of the present embodiment can also be realized as one function of an image forming apparatus including a scanner, such as a copier or a multifunction peripheral. In this case, the image forming apparatus includes a computer system as shown in FIG. Then, for example, the CPU 101 in the image forming apparatus uses the RAM 102 as a work area and executes a predetermined program stored in the ROM 103, the HDD 104, or the like, so that the input image Im0 acquired via the scanner or the network is highly compressed. The function of generating the PDF file FIm is realized. In a series of processes from generation of the input image Im0 to generation of the high-compression PDF file FIm, for a part particularly suitable for hardware implementation, its function is performed by dedicated hardware such as an ASIC (Application Specific Integrated Circuit). It may be a configuration realized by using.

  FIG. 3 is a block diagram illustrating a functional configuration example of the image processing apparatus 1 according to the present embodiment. The image processing apparatus 1 includes, as functional components for generating the high-compression PDF file FIm from the input image Im0, for example, as illustrated in FIG. Unit 14 (an embodiment of a “calculating unit”), a deciding unit 15 (an embodiment of a “deciding unit”), a compression processing unit 16 (an embodiment of a “compression processing unit”), and a file generation unit 17 (an embodiment of a “file Generation means ”. In the image processing apparatus 1 of the present embodiment, an input image Im0 acquired as a processing target is input to the detection unit 10, the calculation unit 14, and the compression processing unit 16, and a high-compression PDF file FIm corresponding to the input image Im0 is stored in a file. Output from the generation unit 17.

  The detection unit 10 is a functional module that detects a line drawing candidate from the input image Im0 to be processed. FIG. 4 is a block diagram illustrating a specific example of the detection unit 10. For example, as illustrated in FIG. 4, the detection unit 10 includes a first detection unit 11 (an embodiment of “first detection unit”), an edge enhancement unit 12, and a second detection unit 13 (“second detection unit”). Detection means ”is included as a submodule.

  The first detection unit 11 performs a process of detecting an edge of a line drawing candidate for the input image Im0 to be processed, and outputs a result (hereinafter, referred to as “edge detection result”). The processing performed by the first detection unit 11 uses, for example, the continuity and pattern of black pixels and white pixels obtained by ternarization of the input image Im0, as in the method described in Patent Literature 2, for example. This is a process of detecting edges forming a line image candidate by separating a line image such as a character from a halftone dot. Here, the edge detection result is, for example, coordinate data representing a coordinate position in the input image Im0 of a pixel group detected as an edge of a line drawing candidate by the first detection unit 11. This edge detection result is input to the edge enhancement unit 12.

  The edge emphasis unit 12 performs a process of emphasizing the edge of the line image candidate included in the input image Im0 to be processed using the edge detection result received from the first detection unit 11, and the edge of the line image candidate is emphasized. Generate an edge enhanced image. Since the edge enhancement process by the edge enhancement unit 12 is performed using the edge detection result of the first detection unit 11, there is little possibility that the edges of the pattern included in the input image Im0 will be enhanced. Therefore, the degree of edge enhancement can be increased, and an edge-enhanced image in which the edges of line drawing candidates are clearly enhanced can be generated. The edge enhancement image generated by the edge enhancement unit 12 is input to the second detection unit 13.

  The second detection unit 13 performs a process of detecting a line drawing candidate on the edge enhanced image received from the edge enhancement unit 12, and outputs a result (hereinafter, referred to as a “candidate detection result”). The processing performed by the second detection unit 13 includes, for example, extracting a connected component of black pixels and white pixels from a binarized image obtained by binarizing the edge-emphasized image, and determining the size of a circumscribed rectangle of the connected component. This is a process of detecting a line drawing candidate based on this. That is, in the method described in Patent Literature 2, the processing up to before the extraction of the character line corresponds to an example of the processing by the second detection unit 13. Here, the candidate detection result is, for example, coordinate data indicating a coordinate position in the input image Im0 of a pixel group detected as a line drawing candidate by the second detection unit 13. This candidate detection result is input to the calculation unit 14 and the determination unit 15 as a detection result of the detection unit 10.

  The calculation unit 14 specifies the positions of all the line drawing candidates included in the input image Im0 to be processed using the candidate detection results received from the detection unit 10, and calculates the number of colors of each line drawing candidate and the background color of each line drawing candidate. And the color of each line drawing candidate is calculated by, for example, the method described above. The calculation unit 14 also calculates the aspect ratio of the circumscribed rectangle of each line drawing candidate, the line width (character thickness) of each line drawing candidate, and the like. The aspect ratio of the circumscribed rectangle of the line drawing candidate can be calculated from the number of pixels arranged in the vertical direction and the number of pixels arranged in the horizontal direction of the circumscribed rectangle. The line width of a line drawing candidate can be calculated, for example, from the distance (number of pixels) between edges of the line drawing candidate. Also, the line width of the line drawing candidate (whether it is a bold character) may be calculated from the ratio of the number of pixels of the line drawing region to the total number of pixels of the circumscribed rectangle of the line drawing candidate. The calculation results of the number of colors of each line drawing candidate, the background color of each line drawing candidate, the color of each line drawing candidate, the aspect ratio of the circumscribed rectangle of each line drawing candidate, and the line width of each line drawing candidate by the calculation unit 14 are sent to the determination unit 15. Is entered.

  Using the calculation result received from the calculation unit 14, the determination unit 15 determines whether or not each of the line image candidates detected by the detection unit 10 is a line image to be subjected to the above-described first compression process. That is, the determination unit 15 performs the detection based on the number of colors of the line drawing candidate calculated by the calculation unit 14, the background color of the line drawing candidate, the color of the line drawing candidate, the aspect ratio of the circumscribed rectangle of the line drawing candidate, the line width of the line drawing candidate, and the like. It is determined whether the line drawing candidate detected by the unit 10 satisfies a predetermined condition. Then, only line drawing candidates satisfying a predetermined condition are determined as line drawings to be subjected to the first compression processing. Hereinafter, specific examples of the conditions used by the determination unit 15 will be described. However, the conditions illustrated here are merely examples, and the present invention is not limited to these examples.

  Condition 1: The background color of the line drawing candidate is classified as white (or the background color of the document), the number of colors of the line drawing candidate is 1, and the aspect ratio of the circumscribed rectangle of the line drawing candidate is in the range of 1/3 to 3. In this case, the line drawing candidate is determined as a line drawing. This is because such a line drawing candidate that satisfies the condition 1 is likely to be a character suitable for the first compression processing. An upper limit and a lower limit are set for the size of the line drawing candidate, and even if the size of the line drawing candidate which satisfies the condition 1 is not less than the upper limit or not more than the lower limit, it may not be determined as a line drawing.

  Condition 2: The background color of the line drawing candidate is classified as white (or the background color of the document), the number of color of the line drawing candidate is 2 or more, and the line drawing candidate is classified into black and red or blue. If the aspect ratio of the circumscribed rectangle of the line drawing candidate is close to 1, the line drawing candidate is determined as a line drawing. A line drawing candidate that satisfies condition 2 is more likely to be treated as a line drawing, such as a portion where a red character or a blue character overlaps a black character, or a portion where a seal is stamped on a gray “mark” character. This is because it is highly possible that the object can realize image quality. An upper limit and a lower limit are set for the size of the line drawing candidate, and even if the size of the line drawing candidate that satisfies the condition 2 is not less than the upper limit or not more than the lower limit, it may not be determined as a line drawing. Also, when an object of two or more colors is treated as one line drawing as in this example, when generating the second image layer Im2 expressing the color of the line drawing, a chromatic color is brought to the front rather than black. It is desirable to express

  Condition 3: The background color of the line drawing candidate is classified into a color other than white (or the background color of the document), the number of colors of the line drawing candidate is 1, the color of the line drawing candidate and the background color are classified into different hues, and If the aspect ratio of the circumscribed rectangle is in the range of 1/3 to 3, the line image candidate is determined as a line image. This is because a line drawing candidate that satisfies the condition 3 is likely to be a character on a color ground suitable for the first compression processing. An upper limit and a lower limit are set for the size of the line drawing candidate, and even if the size of the line drawing candidate that satisfies the condition 3 is not less than the upper limit or not more than the lower limit, it may not be determined as a line drawing. The upper limit and the lower limit of the size of the line drawing candidate may be changed according to the required compression ratio. If the color of the line drawing candidate is an achromatic color, a modification may be made such that the difference in hue from the background color is not considered, or the number of colors of the line drawing candidate is regarded as one.

  Condition 4: The number of colors of the line drawing candidate is 1, the aspect ratio of the line drawing candidate is in the range of 1/3 to 3, the line width of the line drawing candidate is equal to or more than a predetermined value, and the color variance of the line drawing candidate is predetermined. If it is less than the value, the line drawing candidate is determined as a line drawing. This is because a line drawing candidate that satisfies the condition 4 is more likely to be a thick solid character with high compression and high image quality when the first compression processing is performed. Note that a lower limit may be set for the size of the line drawing candidate, and even if the size of the line drawing candidate that satisfies the condition 4 is equal to or smaller than the lower limit, it may not be determined as a line drawing.

  The determination unit 15 determines a line image to be subjected to the first compression process among the line image candidates detected by the second detection unit 13 by using, for example, at least one of the conditions 1 to 4 described above. Outputs line drawing data related to line drawing. The line drawing data includes coordinate data representing the coordinate position of the pixel group constituting the line drawing in the input image Im0, and the color of the line drawing (the color of the line drawing candidate calculated by the calculation unit 14 before being determined as the line drawing). included. This line drawing data is input to the compression processing unit 16.

  The compression processing unit 16 uses the line drawing data received from the determination unit 15 to convert the input image Im0 to be processed into a first image layer Im1 that is a binary image including only line drawings and a multi-color image representing the color of the line drawing. A second image layer Im2, which is a value image, and a third image layer Im3, which is a multi-value image expressing a pattern other than a line drawing and a background, are generated. Then, the compression processing unit 16 performs a first compression process suitable for compressing a line image on the first image layer Im1 and the second image layer Im2, and performs a first compression process on the third image layer Im3. Performs a second compression process suitable for compressing a picture or background. The method of compression is not particularly limited. A method suitable for compressing a line image is applied to the first image layer Im1 and the second image layer Im2, and a pattern is applied to the third image layer Im3. The compression processing may be performed by a method suitable for compression of the background or the background. The first image layer Im1, the second image layer Im2, and the third image layer compressed by the compression processing unit 16 are input to the file generation unit 17.

  The file generation unit 17 integrates the first image layer Im1, the second image layer Im2, and the third image layer Im3 received from the compression processing unit 16 on one image file in the PDF format, and forms the input image Im0. Generate a corresponding high-compression PDF file FIm. The format of the image file that integrates the first image layer Im1, the second image layer Im2, and the third image layer Im3 is not limited to the PDF format. For example, a plurality of image layers such as the JPM format may be used. Various formats can be used that are superimposed into one image. As described above, the high-compression PDF file FIm generated by the file generation unit 17 is stored in, for example, the HDD 104 or transmitted to a host computer connected to a network via the network I / F 105.

<Operation of image processing device>
Next, an outline of the operation of the image processing apparatus 1 according to the present embodiment configured as described above will be described. FIG. 5 is a flowchart illustrating the flow of the operation of the image processing apparatus 1 according to the present embodiment.

  When the operation of the image processing apparatus 1 according to the present embodiment is started, first, in step S1, an input image Im0 to be processed is obtained. The input image Im0 is input to the detection unit 10, the calculation unit 14, and the compression processing unit 16.

  Next, in step S2, a process by the first detection unit 11 of the detection unit 10 is performed on the input image Im0 obtained in step S1. Then, the edge detection result output from the first detection unit 11 is input to the edge enhancement unit 12.

  Next, in step S3, the edge emphasizing unit 12 performs a process of emphasizing the edge of the line drawing candidate included in the input image Im0 acquired in step S1 using the edge detection result of step S2, and Generated. This edge-enhanced image is input to the second detection unit 13.

  Next, in step S4, the processing by the second detection unit 13 is performed on the edge-enhanced image generated in step S3. Then, the candidate detection result output by the second detection unit 13 is input to the calculation unit 14 and the determination unit 15.

  Next, in step S5, the calculation unit 14 uses the candidate detection result in step S4 to determine the number of colors of the line image candidate and the background of the line image candidate for each line image candidate included in the input image Im0 acquired in step S1. Processing for calculating the color, the color of the line drawing candidate, the aspect ratio of the circumscribed rectangle of the line drawing candidate, the line width of the line drawing candidate, and the like are performed. Then, the calculation result by the calculation unit 14 is input to the determination unit 15.

  Next, in step S6, the determination unit 15 uses the calculation result in step S5 to determine whether each of the line image candidates in the input image Im0 indicated by the candidate detection result in step S4 is a line image to be subjected to the first compression process. A process to determine whether or not to do so is performed. Then, line drawing data relating to the line drawing to be subjected to the first compression processing is input to the compression processing unit 16.

  Next, in step S7, the first image layer Im1, the second image layer Im2, and the third image layer Im3 are processed by the compression processing unit 16 from the input image Im0 to be processed using the line drawing data in step S6. Generated and individual compression processing is performed on each image layer. The compression processing performed on the first image layer Im1 and the second image layer Im2 is a first compression processing suitable for compressing a line drawing. The compression process performed on the third image layer Im3 is a second compression process suitable for compressing a picture or a background.

  Next, in step S8, the first image layer Im1, the second image layer Im2, and the third image layer Im3 subjected to the compression processing in step S7 are integrated into one image file by the file generation unit 17. The process is performed, and a high-compression PDF file FIm corresponding to the input image Im0 acquired in step S1 is generated.

  Finally, in step S9, the high-compression PDF file FIm generated in step S8 is output from the file generation unit 17 and stored in, for example, the HDD 104, or the host connected to the network via the network I / F 105. Sent to a computer, etc.

<Specific example of processing by first detection unit>
Next, a specific example of the process performed by the first detection unit 11 will be described with reference to FIG. As described above, the processing performed by the first detection unit 11 separates a line image such as a character from a halftone dot using the continuity and pattern of black and white pixels obtained by ternarization of the input image Im0. By doing so, it is a process of detecting an edge constituting a line drawing candidate. FIG. 6 is a flowchart illustrating an example of a process performed by the first detection unit 11.

  The first detection unit 11 first performs MTF correction by filtering on the input image Im0 to be processed (step S101). The MTF correction is a pre-processing performed to improve the accuracy of the next ternarization. As the filter for MTF correction, for example, a filter described in Patent Document 2 can be used.

  Next, the first detection unit 11 performs ternarization on the MTF-corrected input image Im0 (Step S102), and divides the input image Im0 into black pixels, white pixels, and gray pixels.

  Next, as described in, for example, Patent Document 2, the first detection unit 11 extracts a black line image or a white line image from the input image Im0 by performing pattern matching of black connected pixels or white connected pixels, thereby performing input. An edge of a line drawing candidate included in the image Im0 is detected (Step S103).

  Then, the first detection unit 11 outputs the processing result of step S103 as an edge detection result (step S104), and ends a series of processing.

<Specific example of processing by second detection unit>
Next, a specific example of the process performed by the second detection unit 13 will be described with reference to FIG. As described above, the process performed by the second detection unit 13 extracts the connected components of black pixels and white pixels from the binarized image obtained by binarizing the edge-emphasized image, and calculates the size of the circumscribed rectangle of the connected components. This is a process of detecting a line drawing candidate based on the above. FIG. 7 is a flowchart illustrating an example of a process performed by the second detection unit 13.

  The second detection unit 13 first performs binarization on the edge-enhanced image generated by the edge enhancement unit 12 (Step S201). The binarization here is a process for extracting an object having lower luminance than the background, and an appropriate threshold value that can distinguish the object having low luminance from the background is set. In addition, a dynamic threshold binarization may be used to increase the separation accuracy from the background.

  Next, as described in Patent Literature 1, for example, the second detection unit 13 executes a run of black pixels arranged in the horizontal direction and a black pixel arranged in the vertical direction from the binarized image obtained in step S201. Are connected to obtain a connected component (step S202). Then, the second detection unit 13 detects, as a line drawing candidate, a connected component that can be distinguished from the picture, based on the size of the circumscribed rectangle of the connected component, among the connected components acquired in step S202 (step S203).

  Next, the second detection unit 13 performs binarization again on the edge enhanced image generated by the edge enhancement unit 12 (Step S204). The binarization here is a process for extracting an object with higher luminance than the background, and an appropriate threshold value that can distinguish the object with high luminance from the background is set. In addition, a dynamic threshold binarization may be used to increase the separation accuracy from the background.

  Next, as described in Patent Literature 1, for example, the second detection unit 13 executes a run of white pixels arranged in the horizontal direction and a white pixel arranged in the vertical direction from the binarized image obtained in step S204. Are connected to obtain a connected component (step S205). Then, the second detection unit 13 detects, as a line drawing candidate, a connected component that can be distinguished from a picture, based on the size of a circumscribed rectangle of the connected component, among the connected components acquired in step S205 (step S206).

  Next, the second detection unit 13 determines whether there is a line drawing candidate whose circumscribed rectangle overlaps with the line drawing candidate detected in step S203 and the line drawing candidate detected in step S206 (step S207). If there are line drawing candidates with overlapping circumscribed rectangles (step S207: Yes), the sizes of the circumscribed rectangles of those line image candidates are compared, and the line image candidate with the smaller circumscribed rectangle is deleted (step S208). Then, the second detection unit 13 outputs a candidate detection result as a final line drawing candidate, of the line drawing candidates detected in step S203 and step S206, which remain without being deleted in step S208 (step S209). ), And a series of processing ends.

<Specific example of processing by calculation unit>
Next, a specific example of the processing by the calculation unit 14 will be described with reference to FIG. FIG. 8 is a flowchart illustrating an example of a process performed by the calculation unit 14.

  The calculation unit 14 first extracts one of the line drawing candidates included in the input image Im0 to be processed based on the candidate detection result received from the detection unit 10 (Step S301).

  Next, the calculation unit 14 selects an area having a predetermined size and shape adjacent to the line drawing candidate extracted in step S301 (there may be a gap of about one pixel), and the RGB of each pixel in this area is selected. The data is converted into HSV data, and the average value of the HSV data of each pixel converted from the RGB data is calculated as the background color of the line drawing candidate (step S302).

  Next, the calculation unit 14 converts the RGB data of each pixel constituting the line drawing candidate extracted in step S301 into HSV data, and uses the HSV data of each pixel converted from the RGB data, for example, by the above-described method. The number of candidate colors is calculated (step S303). Further, the calculation unit 14 calculates the color of the line drawing candidate by using the HSV data of each pixel constituting the line drawing candidate, for example, by the above-described method (step S304). At this time, the calculation unit 14 excludes pixels that are close to the background color of the line drawing candidate calculated in step S302 (for example, pixels whose Euclidean distance in the HSV color space is within a predetermined value) from among the pixels constituting the line drawing candidate. The number of colors or the color of a line drawing candidate may be calculated.

  Next, the calculation unit 14 obtains a circumscribed rectangle of the line drawing candidate extracted in step S301, counts the number of pixels arranged in the vertical direction and the number of pixels arranged in the horizontal direction of the circumscribed rectangle, and calculates the length and width of the circumscribed rectangle of the line image candidate. The ratio is calculated (step S305). Further, the calculation unit 14 calculates the line width of the line drawing candidate from the distance between the edges (the number of pixels) of the line drawing candidate extracted in step S301 (step S306).

  Next, the calculation unit 14 determines whether or not there is an unprocessed line drawing candidate (step S307). If there is an unprocessed line drawing candidate (step S307: Yes), the calculation unit 14 returns to step S301 and performs the subsequent processing. repeat. On the other hand, when the processing is completed for all the line drawing candidates (Step S307: No), the calculation unit 14 outputs the calculation results of Steps S302 to S306 for each line drawing candidate (Step S308), and ends a series of processing. I do.

<Specific example of processing by the determination unit>
Next, a specific example of the process performed by the determining unit 15 will be described with reference to FIG. FIG. 9 is a flowchart illustrating an example of a process performed by the determining unit 15.

  The determining unit 15 first extracts one of the line drawing candidates included in the input image Im0 to be processed based on the candidate detection result received from the detecting unit 10 (Step S401).

  Next, the determination unit 15 determines whether the line drawing candidate extracted in step S401 satisfies any of the above-described conditions 1 to 4 using the calculation results received from the calculation unit 14 (steps S402 to S402). Step S405). If the line drawing candidate extracted in step S401 satisfies any of the above conditions 1 to 4 (step S402: Yes, step S403: Yes, or step S404: Yes, or step S405: Yes), the determination unit The step 15 determines the line drawing candidate extracted in step S401 as a line drawing (step S406). If none of Condition 1 to Condition 4 described above is satisfied (Step S402: No, Step S403: No, Step S404: No, and Step S405: No), the line drawing candidate extracted in Step S401 is Discarded.

  Next, the determination unit 15 determines whether there is an unprocessed line drawing candidate (step S407), and if there is an unprocessed line drawing candidate (step S407: Yes), returns to step S401 to perform the subsequent processing. repeat. On the other hand, when the processing is completed for all the line drawing candidates (Step S407: No), the determination unit 15 outputs the line drawing data regarding the line drawing determined in Step S406 (Step S408), and ends a series of processing.

<Specific example of processing by compression processing section>
Next, a specific example of the processing by the compression processing unit 16 will be described with reference to FIG. FIG. 10 is a flowchart illustrating an example of a process performed by the compression processing unit 16.

  The compression processing unit 16 first extracts a line drawing included in the input image Im0 to be processed using the line drawing data received from the determination unit 15, and the first image layer Im1 that is a binary image including only the line drawing. Then, a second image layer Im2, which is a multi-valued image expressing the color of the line drawing, and a third image layer Im3, which is a multi-valued image expressing a pattern and background other than the line drawing, are generated (step S501).

  Next, the compression processing unit 16 applies a compression process (for example, a compression process using a coding method such as an MMR to a binary image) suitable for compressing a line image on the first image layer Im1 generated in step S501. First compression processing) is performed, and the compressed first image layer Im1 is output (step S502). Further, the compression processing unit 16 encodes the second image layer Im2 generated in step S501, for example, into a multi-valued image, such as JPEG, which has a lower resolution than the compression processing of the third image layer Im3. A compression process (first compression process) suitable for compressing a line image, such as a compression process by a method, is performed, and a compressed second image layer Im2 is output (step S503). Further, the compression processing unit 16 encodes the third image layer Im3 generated in step S501, for example, into a multi-valued image such as JPEG, which has a higher resolution than the compression processing of the second image layer Im2. A compression process (second compression process) suitable for compressing a picture or a background, such as a compression process by a method, is performed, and a compressed third image layer Im3 is output (step S504).

  The compressed first image layer Im1, second image layer Im2, and third image layer Im3 output from the compression processing unit 16 are output to the file generation unit 17 on one image file in, for example, a PDF format. By the integration, a high-compression PDF file FIm corresponding to the input image Im0 is generated.

<Effects of Embodiment>
As described above in detail with a specific example, the image processing apparatus 1 of the present embodiment detects a line drawing candidate from the input image Im0 to be processed, and detects the number of colors of the detected line drawing candidate, the background color, The color, the aspect ratio of the circumscribed rectangle, the line width, and the like are calculated. Then, by comparing the calculated number of colors of the line drawing candidates, the background color, the color, the aspect ratio of the circumscribed rectangle, the line width, etc. with predetermined conditions, it is determined whether or not the detected line drawing candidate is a line drawing. , A first compression process suitable for compressing a line drawing is performed, and a second compression process is performed on a pattern or background other than a line drawing. Therefore, according to the image processing apparatus 1 of the present embodiment, an object that is not suitable for the first compression processing can be detected as a line drawing from the input image Im0, or an object that is suitable for the first compression processing can be detected as a line drawing. It is possible to effectively suppress the inconvenience, and to improve the efficiency of compression of the input image Im0.

<First Modification>
In the embodiment described above, among the line drawing candidates detected by the detection unit 10, a line drawing candidate that satisfies any of the above-described conditions 1 to 4 is determined as a line drawing. The conditions to be fixed are not limited to this. For example, in addition to the above-described conditions 1 to 4, when a line drawing candidate detected by the detection unit 10 forms a character line, a condition that the line drawing candidate is determined as a line drawing may be added.

  Whether or not the line drawing candidate detected by the detection unit 10 forms a character line can be determined by, for example, a method described in Patent Document 1. In the present modification, for example, it is determined whether or not a character line is formed for each of the line drawing candidates detected by the detection unit 10, and the line drawing candidates forming the character line are determined as line drawings. Then, for the remaining line drawing candidates that are determined not to form a character line, the determination based on the above-described conditions 1 to 4 is performed. Therefore, according to the present modification, a fragmentary character or the like that cannot be detected because the character line is not formed by the method described in Patent Document 1 can be appropriately detected as a line drawing.

<Second modification>
The above-described conditions 1 to 4 are merely examples, and various modifications are possible. For example, among the line drawing candidates detected by the detection unit 10, a line drawing candidate having one or two or less colors is determined as a line drawing, and the line drawing candidate is determined as a line drawing based on only the number of colors of the line drawing candidate. It may be determined whether or not. Further, for example, when the background color of the line drawing candidate detected by the detection unit 10 is white or the background color of the document, and the number of color of the line drawing candidate is 1 or 2 or less, the line drawing candidate is regarded as a line drawing. Alternatively, it may be determined whether or not the line drawing candidate is a line drawing based on the number of colors of the line drawing candidate and the background color.

<Third Modification>
In the above-described embodiment, the edge detection result by the first detection unit 11 is used for generating an edge-enhanced image. However, the edge detection result by the first detection unit 11 is used for detecting a line drawing candidate. May be. For example, by performing an OR operation on the edge detection result by the first detection unit 11 and the candidate detection result by the second detection unit 13, a region having an edge detected by the first detection unit 11 as a contour is determined as a line drawing candidate. You may add it. Further, in this case, among the regions having the outlines of the edges detected by the first detection unit 11, those excluding the line drawing on the color ground and the white line drawing may be added to the line drawing candidates. Whether the line drawing is a line drawing on a color ground or a white line drawing may be determined by calculating the color of the line drawing or the background color by the same method as the calculation unit 14, or based on the color calculated by a simpler method. May be determined. In addition, the second detection unit 13 detects a line image candidate from the input image Im0 without generating an edge-enhanced image, and detects the edge detection result by the first detection unit 11 and the candidate by the second detection unit 13. A configuration may be such that a line drawing candidate is detected from the input image Im0 by an OR operation with the detection result.

<Fourth modification>
The method for detecting a line drawing candidate described in the above-described embodiment is an example, and the present invention is not limited to this. Any of the existing methods for detecting a line drawing such as a character from an image can be used for detecting a line drawing candidate.

<Supplementary explanation>
The functional components of the image processing apparatus 1 (the detection unit 10 (the first detection unit 11, the edge enhancement unit 12, and the second detection unit 13), the calculation unit 14, and the determination described in the above-described embodiment and the modification example) The unit 15, the compression processing unit 16, and the file generation unit 17) can be realized, for example, by cooperation of the hardware and software (program) shown in FIG. 2 as described above. In this case, the program is recorded in a computer-readable recording medium such as a CD-ROM, a flexible disk (FD), a CD-R, and a DVD in a file that can be installed or executed in the image processing apparatus 1. Provided. Further, the program may be stored on a computer connected to a network such as the Internet, and provided by being downloaded to the image processing apparatus 1 via the network. Further, the program may be provided or distributed via a network such as the Internet. Further, the above-described program may be provided by being incorporated in advance in, for example, the ROM 103 or the HDD 104 in the image processing apparatus 1.

  In addition, some or all of the functional components of the image processing apparatus 1 described in the above-described embodiments and modified examples may be partially or entirely replaced with, for example, an ASIC (Application Specific Integrated Circuit) or an FPGA (Field-Programmable Gate Array). It can also be realized using dedicated hardware.

  Further, in the above-described embodiment and modified examples, an example in which the image processing apparatus 1 is realized as a single apparatus is assumed. However, functional components of the image processing apparatus 1 are distributed to a plurality of physically separated apparatuses. The operation as the image processing apparatus 1 may be realized by cooperation of the plurality of apparatuses.

  As described above, the specific embodiments and the modified examples of the present invention have been described. However, the above-described embodiments show one application example of the present invention. The present invention is not limited to the above-described embodiments as they are, and can be embodied in the implementation stage by adding various modifications and changes without departing from the gist thereof.

REFERENCE SIGNS LIST 1 image processing device 10 detection unit 11 first detection unit 12 edge enhancement unit 13 second detection unit 14 calculation unit 15 determination unit 16 compression processing unit 17 file generation unit Im0 input image FIm high compression PDF file

Japanese Patent No. 4471202 Japanese Patent No. 3088010

Claims (10)

Detecting means for detecting a candidate of a target area to be subjected to the first compression processing from the image;
For each of the detected candidates, calculating means for calculating the number of colors representing the number of colors of the candidates,
Determining means for determining whether to set the candidate as the target area based on at least the number of colors of the candidate;
Compression processing means for performing the first compression processing on the target area and performing a second compression processing different from the first compression processing on an area other than the target area included in the image; and, with a,
The calculating means further calculates, for each of the detected candidates, a background color that is a color of an area adjacent to the candidate, and a color of the candidate,
The image processing device , wherein the determination unit determines whether or not the candidate is the target area based on the number of colors of the candidate, the background color of the candidate, and the color of the candidate . The calculating unit further calculates, for each of the detected candidates, an aspect ratio of a circumscribed rectangle of the candidate,
The image processing according to claim 1 , wherein the determining unit determines whether or not the candidate is set as the target area based on at least a number of colors of the candidate and an aspect ratio of a circumscribed rectangle of the candidate. apparatus. Detecting means for detecting a candidate of a target area to be subjected to the first compression processing from the image;
For each of the detected candidates, calculating means for calculating the number of colors representing the number of colors of the candidates,
Determining means for determining whether to set the candidate as the target area based on at least the number of colors of the candidate;
Compression processing means for performing the first compression processing on the target area and performing a second compression processing different from the first compression processing on an area other than the target area included in the image; And
The calculating means further calculates, for each of the candidates, a line width of the candidate,
Said determining means includes at least a color number of the candidate, based on the line width of the candidate to determine whether the candidate to the target area, images processing device. The calculating means calculates an average value of the color data of each pixel constituting the candidate, and if a ratio of pixels having color data whose difference from the average value is equal to or more than a predetermined value is less than a predetermined value, the candidate is calculated. to number of colors and 1, the image processing apparatus according to any one of claims 1 to 3. The detection means,
First detection means for detecting the candidate edge from the image;
And a second detecting means for detecting the candidate detection result from the image edges of the candidates is emphasized by using the first detecting means, according to any one of claims 1 to 4 Image processing device. File generation for integrating the target area subjected to the first compression processing and an area other than the target area subjected to the second compression processing to generate a compressed image file corresponding to the image further comprising means, image processing apparatus according to any one of claims 1 to 5. Detecting a candidate of a target area to be subjected to the first compression processing from the image;
For each of the detected candidates, calculating the number of colors representing the number of colors of the candidates,
Determining whether to set the candidate as the target area based on at least the number of colors of the candidate;
Performing the first compression processing on the target area and performing a second compression processing different from the first compression processing on an area other than the target area included in the image; only including,
In the calculating step, for each of the detected candidates, further calculate a background color that is a color of an area adjacent to the candidate, and a color of the candidate,
The image processing method, wherein in the determining step, whether or not the candidate is set as the target area is determined based on the number of colors of the candidate, the background color of the candidate, and the color of the candidate . Detecting a candidate of a target area to be subjected to the first compression processing from the image;
For each of the detected candidates, calculating the number of colors representing the number of colors of the candidates,
Determining whether to set the candidate as the target area based on at least the number of colors of the candidate;
Performing the first compression processing on the target area and performing a second compression processing different from the first compression processing on an area other than the target area included in the image; Including
In the calculating step, for each of the candidates, further calculating the line width of the candidates,
The image processing method, wherein in the determining, based on at least the number of colors of the candidate and the line width of the candidate, it is determined whether or not the candidate is to be the target area. On the computer,
A function of detecting a candidate of a target area to be subjected to the first compression processing from the image;
For each of the detected candidates, a function of calculating the number of colors representing the number of colors of the candidates,
A function of determining whether to set the candidate as the target area based on at least the number of colors of the candidate;
A function of performing the first compression process on the target region and performing a second compression process different from the first compression process on a region other than the target region included in the image; to achieve,
The calculating function further calculates, for each of the detected candidates, a background color that is a color of an area adjacent to the candidate, and a color of the candidate,
The determining function is a program for determining, based on the number of colors of the candidate, the background color of the candidate, and the color of the candidate, whether or not the candidate is the target area . On the computer,
A function of detecting a candidate of a target area to be subjected to the first compression processing from the image;
For each of the detected candidates, a function of calculating the number of colors representing the number of colors of the candidates,
A function of determining whether or not to set the candidate as the target area based on at least the number of colors of the candidate;
A function of performing the first compression process on the target region and performing a second compression process different from the first compression process on a region other than the target region included in the image; To realize
The calculating function further calculates, for each of the candidates, a line width of the candidate,
A program for determining whether or not to set the candidate as the target area based on at least the number of colors of the candidate and the line width of the candidate.

Images