JP2008205994A - Image encoding apparatus, image decoding apparatus, and image processing system and program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an image encoding apparatus capable of preventing image quality from being deteriorated when inputting a non-deteriorated image. <P>SOLUTION: In an image encoding program 22, a first mark comparing unit 220 and a deterioration determining unit 224 acquire the deterioration state of a binary input image on the basis of a first pattern matching technique. In such a case, based on the statistic information of first pattern matching, the deterioration determining unit 224 determines whether the input image is a deteriorated image or a non-deteriorated image. Based on a second pattern matching technique corresponding to the acquired deterioration state, a second mark comparing unit 202 generates an image pattern dictionary from an image pattern contained in the input image. An encoding unit 206 encodes the generated image pattern dictionary and a position where the image pattern contained in the relevant image pattern dictionary appears in the input image. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to an image encoding device, an image decoding device, an image processing system, and a program based on pattern matching.

  An image encoding method that extracts a mark such as a character or a figure that repeatedly appears in an input image to create an image dictionary and encodes the appearance position and the dictionary is known as, for example, JBIG2 text encoding processing. Also, when matching the extracted marks with the marks in the dictionary, it is often possible to realize lossy encoding with a high compression rate by reducing the size of the dictionary by allowing a certain degree of difference. Are known.

  However, such an encoding method is not applied to an already deteriorated image (for example, a scanned image or an image on which quantization noise is placed), but applied to an undegraded image such as a CG (Computer Graphics) image. In such a case, image quality deterioration due to replacement may be noticeable. In general, characters included in a CG have characteristics such as equal line width and line spacing, high symmetry and linearity, so that even if a minute change is added to the image, the image deteriorates. It is because it is easy to be recognized that it is doing.

  Patent Document 1 discloses an image reading apparatus having means for extracting a scan quality parameter indicating the quality of a scanned image from image data.

JP 2005-31766 Gazette

  An object of the present invention is to provide an image encoding device capable of preventing deterioration of image quality when a non-degraded image is input.

  In order to achieve the above object, a first image encoding device according to the present invention includes a degradation state accepting unit that accepts an input related to a degradation state of an image, and a pattern corresponding to the degradation state accepted by the degradation state acceptance unit. Based on a matching method, a dictionary generation unit that generates an image pattern dictionary from an image pattern included in a binary input image, an image pattern dictionary generated by the dictionary generation unit, and an image pattern included in the image pattern dictionary Encoding means for encoding the position at which appears in the input image. According to the present invention, when a non-degraded image is input, it is possible to prevent deterioration in image quality.

  The second image encoding device according to the present invention is based on a first pattern matching method, a deterioration state acquisition unit that acquires a deterioration state of a binary input image, and a deterioration acquired by the deterioration state acquisition unit. A dictionary generating unit that generates an image pattern dictionary from an image pattern included in an input image based on a second pattern matching method corresponding to the state; an image pattern dictionary generated by the dictionary generating unit; and the image pattern dictionary Encoding means for encoding the position where the image pattern included in the image appears in the input image. According to the present invention, when a non-degraded image is input, it is possible to prevent deterioration in image quality. When an undegraded image is input, it is possible to prevent deterioration in image quality even if the deterioration state of the image is not input.

  Preferably, the deterioration state acquisition unit determines whether the image pattern included in the input image matches the image pattern to be compared as the first pattern matching. According to the present invention, the amount of calculation required for the first pattern matching can be reduced.

  Preferably, the deterioration state acquisition unit performs pattern matching based on a hash value generated from the image pattern. According to the present invention, the first pattern matching process can be speeded up.

  Preferably, the deterioration state acquisition unit determines the deterioration state of the input image based on a result of pattern matching performed on an image pattern including a number of pixels equal to or greater than a predetermined value. According to the present invention, the result of pattern matching for a small image pattern is excluded from the determination target, so that the accuracy of determination of the deterioration state can be improved.

  Preferably, when the input image is a non-degraded image, the dictionary generation unit performs pattern matching by applying a comparison parameter smaller than that when the input image is a deteriorated image. According to the present invention, it is possible to invalidate lossy encoding according to an input image.

  Preferably, the deterioration state acquisition unit generates a first image pattern dictionary from an image pattern included in the binary input image, and the dictionary generation unit generates the first image pattern generated by the deterioration state acquisition unit. A second image pattern dictionary is generated from the image patterns included in the image pattern dictionary. According to the present invention, since the number of image patterns to be subjected to the second pattern matching process is reduced, the second pattern matching process can be speeded up.

  Preferably, when the input image is an undegraded image, the encoding unit includes a first image pattern dictionary generated by the deterioration state acquiring unit and an image pattern included in the first image pattern dictionary. When the position appearing in the input image is encoded and the input image is a deteriorated image, the second image pattern dictionary generated by the dictionary generation unit and the image pattern included in the second image pattern dictionary are input images. And the position that appears in. According to the present invention, when it is determined that the input image is an undegraded image, the second pattern matching process is omitted, so that the entire image encoding process can be speeded up.

  Preferably, the information processing apparatus further includes deterioration state reception means for receiving an input relating to a deterioration state of the image, and the encoding means is generated by the deterioration state acquisition means when the deterioration state is received by the deterioration state reception means. The first image pattern dictionary and the position where the image pattern included in the first image pattern dictionary appears in the input image are encoded. According to the present invention, even when the degradation state of an input image is known and unknown, the entire image encoding process can be speeded up.

  The first image decoding apparatus according to the present invention is based on a degradation state accepting unit that accepts an input related to a degradation state of an image, and a pattern matching method corresponding to the degradation state accepted by the degradation state accepting unit. A dictionary generation unit that generates an image pattern dictionary from an image pattern included in the input image, an image pattern dictionary generated by the dictionary generation unit, and a position at which the image pattern included in the image pattern dictionary appears in the input image. Code data receiving means for receiving code data encoded by an image encoding device having encoding means for encoding, decoding means for decoding code data received by the code data receiving means, and the decoding And a decoded image generating means for generating a decoded image based on the data decoded by the converting means. ADVANTAGE OF THE INVENTION According to this invention, when decoding the code | cord | chord data with which the non-deteriorated image was encoded, the decoded image by which degradation of the image quality was reduced can be produced | generated.

  The second image decoding apparatus according to the present invention is based on a first pattern matching method, a deterioration state acquisition unit that acquires a deterioration state of a binary input image, and a deterioration acquired by the deterioration state acquisition unit. A dictionary generating unit that generates an image pattern dictionary from an image pattern included in an input image based on a second pattern matching method corresponding to the state; an image pattern dictionary generated by the dictionary generating unit; and the image pattern dictionary Code data receiving means for receiving code data encoded by an image encoding device having an encoding means for encoding a position where an image pattern included in the input image appears in the input image, and received by the code data receiving means Decoding means for decoding the code data, and generating a decoded image based on the data decoded by the decoding means And a decoded image generation unit that. ADVANTAGE OF THE INVENTION According to this invention, when decoding the code | cord | chord data with which the non-deteriorated image was encoded, the decoded image by which degradation of the image quality was reduced can be produced | generated.

  A first image processing system according to the present invention is based on a degradation state accepting unit that accepts an input relating to an image degradation state, and a pattern matching method corresponding to the degradation state accepted by the degradation state acceptance unit. A dictionary generation unit that generates an image pattern dictionary from an image pattern included in the input image, an image pattern dictionary generated by the dictionary generation unit, and a position at which the image pattern included in the image pattern dictionary appears in the input image Encoding means for decoding, decoding means for decoding the code data encoded by the encoding means, and decoded image generation means for generating a decoded image based on the data decoded by the decoding means; Have

  A second image processing system according to the present invention is based on a first pattern matching method, a deterioration state acquisition unit that acquires a deterioration state of a binary input image, and a deterioration state acquired by the deterioration state acquisition unit Based on a second pattern matching method corresponding to the above, a dictionary generating means for generating an image pattern dictionary from an image pattern included in the input image, an image pattern dictionary generated by the dictionary generating means, and an image pattern dictionary Encoding means for encoding a position at which an included image pattern appears in an input image, decoding means for decoding code data encoded by the encoding means, and data decoded by the decoding means And a decoded image generating means for generating a decoded image based on the above.

  The first program according to the present invention is based on a deterioration state receiving step for receiving an input relating to an image deterioration state and a pattern matching method corresponding to the received deterioration state in an image encoding device including a computer. A dictionary generation step of generating an image pattern dictionary from an image pattern included in an input image of values, the generated image pattern dictionary, and a position where the image pattern included in the image pattern dictionary appears in the input image is encoded And causing the image encoding apparatus to execute an encoding step.

  The second program according to the present invention includes a deterioration state acquisition step of acquiring a deterioration state of a binary input image based on a first pattern matching method in an image encoding device including a computer, and the acquired A dictionary generation step of generating an image pattern dictionary from an image pattern included in the input image based on the second pattern matching method corresponding to the deterioration state, the generated image pattern dictionary, and the image pattern dictionary The image encoding apparatus executes an encoding step for encoding a position where an image pattern appears in an input image.

  According to a third program of the present invention, in an image decoding apparatus including a computer, in the image encoding apparatus including the computer, a deterioration state receiving step for receiving an input regarding an image deterioration state and the received deterioration state A dictionary generation step of generating an image pattern dictionary from an image pattern included in a binary input image based on the pattern matching method to be performed, the generated image pattern dictionary, and an image pattern included in the image pattern dictionary are input An encoding step for encoding a position appearing in an image is executed, and an encoded data receiving step for receiving encoded data; a decoding step for decoding the received encoded data; and the decoded data A decoded image generating step for generating a decoded image based on the obtained data Serial causing a computer to execute the picture decoding apparatus.

  The fourth program according to the present invention is an image decoding apparatus including a computer, wherein the image encoding apparatus including the computer acquires a deterioration state of the binary input image based on the first pattern matching method. A dictionary generation step of generating an image pattern dictionary from an image pattern included in an input image based on a state acquisition Sutepu and a second pattern matching method corresponding to the acquired degradation state; and the generated image pattern dictionary A code data receiving step for receiving the encoded data by executing an encoding step for encoding the position where the image pattern included in the image pattern dictionary appears in the input image, and the received code data A decoding step for decoding the image, and a decoded image is generated based on the decoded data. A decoded image generation step of causing a computer to execute the image decoding apparatus.

  According to the present invention, when a non-degraded image is input, it is possible to prevent deterioration in image quality.

First, in order to facilitate understanding of the present invention, the background of the present invention will be described.
In the image encoding process based on pattern matching, marks such as characters and figures are cut out from the input image, and it is determined whether or not this cut out mark and any of the marks in the dictionary are considered to match. . If the clipped mark is considered to match a mark in the dictionary, the clipped mark is replaced with a mark in the dictionary, and if it is not considered to match, the clipped mark is added to the dictionary. When such processing is performed on all the marks included in the input image, the positions of the marks in the dictionary and the input image are encoded.

  FIG. 1 is a diagram illustrating a deteriorated image and a non-degraded image that have been subjected to pattern matching processing, and FIGS. 1A to 1C illustrate the case of a deteriorated image, and FIG. FIG. 1F illustrates the case of an undegraded image. Here, the deteriorated image is quantized into an original image, for example, a read image read by a scanner or the like, or a binary image generated by performing binarization processing on a color or grayscale multi-valued image. This is an image with noise. The non-degraded image is, for example, a CG image generated by predetermined application software.

  As illustrated in FIG. 1A, when the characters “times” and “phrase” are included in the input image, it is considered that the image portion of “mouth” of these characters is matched by the pattern matching process. is there. In this case, the image portion of one “mouth” is replaced with the image portion of the other “mouth”. In this example, as illustrated in FIG. 1B, the “mouth” image portion of the “times” character is replaced with the “mouth” image portion of the “phrase” character.

  When multiple image parts are considered to be matched by the pattern matching process, the image part is averaged, and the multiple image parts are replaced with the averaged image part. There is also. In FIG. 1C, averaging processing is performed on two “mouth” image portions, and the “mouth” image portions of “times” and “phrase” are averaged “mouth” images. The characters replaced with parts are illustrated.

  On the other hand, for example, as shown in FIGS. 1D and 1E, in the non-deteriorated image, the “mouth” image portion of the “times” character is the “mouth” image of the “phrase” character. In the case of replacement with a part, the difference in the distance between the lines, as indicated by the arrows in the figure, becomes conspicuous. Further, for example, as shown in FIG. 1F, in the non-degraded image, when the “mouth” image portion is replaced with the averaged “mouth” image portion, the line indicated by the arrow in the figure. The difference in width is noticeable. Thus, when the input image is a non-degraded image, the image after the pattern matching process and the replacement process is deteriorated compared to the case where the same process is applied to the deteriorated image. Are often recognized.

FIG. 2 is a diagram illustrating an input image for explaining deterioration of image quality in a deteriorated image and a non-degraded image.
As shown in FIGS. 2A and 2B, there is only one dot difference between the two figures. When the input image is a deteriorated image, this difference is recognized as a difference due to deteriorated noise, and it is considered that these two figures may be regarded as the same. On the other hand, when the input image is an undegraded image, the graphic shown in FIG. 2A is recognized as indicating the character image of the alphabet “O” (Oh), and the graphic shown in FIG. Needs to be recognized as indicating a character image of “Q” (queue). That is, the two figures need to be clearly distinguished.

  Hereinafter, an image encoding device according to an embodiment of the present invention, an image decoding device that decodes code data generated by the image encoding device to generate a decoded image, and an image encoding device and an image decoding device are provided. A specific image processing system will be described.

FIG. 3 is a diagram showing a hardware configuration of the image encoding device 2 and the image decoding device 4 according to the embodiment of the present invention, centering on the control device 100.
As shown in FIG. 3, the image encoding device 2 and the image decoding device 4 include a control device 100 including a CPU 102 and a memory 104, and a communication device that transmits and receives data to and from an external computer via a network. 106, a storage device 108 such as a hard disk drive, and a user interface (UI) device 110 including a display device such as a liquid crystal display and a pointing device such as a keyboard and a mouse.

  The image encoding device 2 and the image decoding device 4 are, for example, general-purpose computers in which an image encoding program 20 and an image decoding program 40 described later are installed. The communication device 106, the storage device 108, or the FD, CD, DVD The image data is acquired via the recording medium 112 or the like, and the data is encoded and decoded. Note that the image encoding device 2 may acquire image data optically read by a scanner function of a multifunction peripheral (not shown) or a scanner device.

  In the image encoding device 2, the UI device 110 constitutes a degradation state accepting unit that accepts an input related to a degradation state of an image. Specifically, a screen for accepting whether the image is a deteriorated image or an undegraded image is displayed on the display device, and the input is accepted via the input device. The UI device 110 outputs the received input to the CPU 102. The deterioration state accepting unit may be realized by the communication device 106. In this case, the communication device 106 receives input of data regarding the image degradation state via a network (not shown) and outputs the data to the CPU 102.

FIG. 4 is a diagram illustrating a functional configuration of the image encoding program 20 executed by the control device 100 of the image encoding device 2.
As illustrated in FIG. 4, the image encoding program 20 includes a mark cutout unit 200, a mark comparison unit 202, a mark dictionary 204, an encoding unit 206, and a comparison parameter determination unit 208. For example, the image encoding program 20 is supplied to the control device 100 via the communication device 106, loaded into the memory 104, and specifically uses hardware on an OS (not shown) operating on the control device 100. Executed. The image encoding program 20 may be stored in the recording medium 112 and supplied to the image encoding device 2 or may be supplied via the communication device 106. The following programs are also supplied and executed in the same manner. All or some of the functions of the image encoding program 20 may be realized by hardware such as an ASIC provided in the image encoding device 2.

  In the image encoding program 20, the mark cutout unit 200 receives a binary input image via the communication device 106, the storage device 108, or the recording medium 112, extracts an image pattern included in the input image, and compares the marks. Output to the unit 202. The image pattern is a partial image of a predetermined area included in the input image, and is, for example, a character or a graphic. For example, the mark cutout unit 200 cuts out black pixels connected in the vicinity of 8 from the input image as one image pattern. Hereinafter, the image pattern cut out by the mark cutout unit 200 is also referred to as a mark.

  The mark comparison unit 202 generates a later-described mark dictionary 204 based on a pattern matching method. Specifically, the mark comparison unit 202 inputs the mark cut out by the mark cutout unit 200, reads out the mark included in the mark dictionary 204, and compares the cut out mark with the read out mark. The mark comparison unit 202 determines whether or not the mark dictionary 204 includes a mark that can be considered to match the extracted mark. If included, the mark comparison unit 202 determines that the extracted mark is included in the mark dictionary 204. If the mark is not included, an index (identifier) uniquely identifying the extracted mark is added to the mark, and the mark and the index are added to the mark dictionary 204. The mark comparison unit 202 outputs the mark index and drawing position in the input image to the encoding unit 206.

  For example, the mark comparison unit 202 applies a technique such as global template matching or local template matching to determine whether or not the two marks match. In global template matching, mark matching is determined based on global features such as the difference in the number of pixels when marks are superimposed, and in local template matching, for example, local matching when marks are superimposed is determined. The determination is made based on a local feature amount such as a difference in the number of pixels in the region. Note that the mark comparison unit 202 may perform a match / mismatch determination by, for example, comparing feature amounts such as a mark peripheral length and a mark area.

  The mark comparison unit 202 applies a pattern matching method corresponding to the comparison parameter determined by the comparison parameter determination unit 208 described later. For example, if the difference in the number of pixels when the marks are superimposed is less than the comparison parameter, the mark comparison unit 202 determines that the two marks match, and if the difference in the number of pixels is greater than or equal to the comparison parameter, It is determined that the marks do not match. Therefore, the smaller the comparison parameter, the smaller the probability that it will be determined that the two marks match.

  The mark dictionary 204 stores a plurality of different marks and indexes. The mark dictionary 204 is realized by at least one of the memory 104 and the storage device 108. The mark dictionary is also called an image pattern dictionary.

  The comparison parameter determination unit 208 determines a comparison parameter based on the degradation state of the image received by the UI device 110 and outputs it to the mark comparison unit 202. Specifically, when the image is an undegraded image, the comparison parameter determination unit 208 has a lower probability that the mark comparison unit 202 determines that the two marks match each other than when the image is in a degraded state. The comparison parameter is determined so that

  For example, when the image is a non-degraded image, the comparison parameter determination unit 208 makes the comparison parameter smaller than when the image is in a deteriorated state. Preferably, when the image is an undegraded image, the comparison parameter determination unit 208 considers that the two patterns match when the two patterns completely match. Set the comparison parameter to.

  Therefore, the mark comparison unit 202 constitutes a dictionary generation unit that generates a mark dictionary from the marks included in the binary input image based on the pattern matching method corresponding to the deterioration state received by the UI device 110.

  The encoding unit 206 encodes the mark dictionary 204 generated by the mark comparison unit 202 and the position where the mark included in the mark dictionary 204 appears in the input image, and generates code data. The encoding unit 206 may store the encoded data in the storage device 108 or may transmit the encoded data to an external computer such as the image decoding device 4 described later via the communication device 106. The code data may be stored in the recording medium 112 and supplied to the image decoding device 4 or the like.

FIG. 5 is a diagram illustrating the relationship between the input image, the mark dictionary 204, and the encoding process. FIG. 5 (A) illustrates a binary input image, and FIG. 5 (B) illustrates the input image. The mark dictionary 204 generated based on this is illustrated, and FIG. 5C illustrates information regarding the drawing position of the mark.
As illustrated in FIG. 5A and FIG. 5B, marks included in the input image (in this example, “A” and “B”) are cut out from the input image by the mark cutout unit 200, and different marks. Is stored in the mark dictionary 204 by the mark comparison unit 202, and an index is assigned to each mark. The shape of the mark is encoded by the encoding unit 206 as information regarding the mark included in the mark dictionary 204. As illustrated in FIG. 5C, the mark position in the input image and the mark index are output from the mark comparison unit 202 to the encoding unit 206 and encoded by the encoding unit 206. .

FIG. 6 is a flowchart of the image encoding process (S10) executed by the image encoding program 20.
As shown in FIG. 6, in step 100 (S <b> 100), the comparison parameter determination unit 208 of the image encoding program 20 receives a deterioration state of an image input via the UI device 110. For example, the comparison parameter determination unit 208 receives whether the image is a deteriorated image or an undegraded image.
In step 102 (S102), the comparison parameter determination unit 208 determines a comparison parameter based on the received degradation state of the image.

In step 104 (S104), the mark cutout unit 200 accepts a binary input image.
In step 106 (S106), the mark cutout unit 200 cuts out a mark from the input image.

In step 108 (S108), the mark comparison unit 202 performs pattern matching processing on the mark cut out by the mark cutout unit 200 based on the comparison parameter determined by the comparison parameter determination unit 208.
In step 110 (S110), the mark comparison unit 202 determines whether or not the mark dictionary 204 includes a mark that can be considered to match the clipped mark. The mark comparison unit 202 proceeds to the process of S112 if a mark that can be regarded as matching is included, and proceeds to the process of S114 otherwise.

In step 112 (S112), the mark comparison unit 202 replaces the cut mark with a mark that is considered to match the mark.
In step 114 (S114), the mark comparison unit 202 adds the cut mark to the mark dictionary 204, and assigns an index to the mark.

In step 116 (S116), the mark cutout unit 200 determines whether or not all the marks included in the input image have been cut out. The mark cutout unit 200 proceeds to the process of S118 when all the marks have been cut out, and returns to the process of S106 otherwise.
In step 118 (S118), the encoding unit 206 encodes the mark, the mark index, and the position at which the mark appears in the input image included in the mark dictionary 204, and generates code data. The encoding unit 206 outputs the generated code data to the storage device 108, the communication device 106, and the like.

FIG. 7 is a diagram illustrating a functional configuration of the image decoding program 40 executed by the control device 100 of the image decoding device 4.
As illustrated in FIG. 7, the image decoding program 40 includes a code data receiving unit 400, a decoding processing unit 402, and a decoded image generating unit 404. With such a configuration, the image decoding program 40 receives code data encoded by the above-described image encoding device 2 and decodes it to generate a decoded image.

In the image decoding program 40, the code data receiving unit 400 receives code data encoded by the image encoding device 2 and outputs the code data to the decoding processing unit 402.
The decoding processing unit 402 decodes the code data received by the code data receiving unit 400, and outputs the mark included in the mark dictionary 204, the mark index, and the drawing position to the decoded image generation unit 404.

  The decoded image generation unit 404 generates a decoded image based on the information decoded by the decoding processing unit 402. Specifically, the decoded image generation unit 404 generates a decoded image by placing a mark identified by the index at the drawing position based on the index and the drawing position of the mark.

FIG. 8 is a flowchart of the image decoding process (S20) by the image decoding program 40.
As shown in FIG. 8, in step 200 (S200), the code data receiving unit 400 of the image decoding program 40 receives the encoded code data.

In step 202 (S202), the decoding processing unit 402 decodes the received code data.
In step 204 (S204), the decoded image generation unit 404 generates a decoded image based on the decoded data.
In step 206 (S206), the decoded image generation unit 404 outputs the generated decoded image.

  The image processing system according to the embodiment of the present invention includes the image encoding device 2 and the image decoding device 4 described above. In the image processing system, the image encoding device 2 and the image decoding device 4 are connected via a network or the like (not shown). Code data and the like are transmitted and received between the image encoding device 2 and the image decoding device 4 via the network by the communication device 106 such as the image encoding device 2.

  Note that the image encoding device 2 and the image decoding device 4 may be included in the same casing. Further, the image encoding program 20 and the image decoding program 40 may operate on the same control device 100.

Next, an image encoding device 2 according to the second embodiment of the present invention will be described.
The image encoding device 2 according to the present embodiment is different from the image encoding device 2 according to the first embodiment in that a deterioration state of the input image is determined from a binary input image.

FIG. 9 is a diagram illustrating a functional configuration of the image encoding program 22 executed by the control device 100 of the image encoding device 2 according to the present embodiment.
As shown in FIG. 9, the image encoding program 22 includes a mark cutout unit 200, a first mark comparison unit 220, a first mark dictionary 222, a deterioration determination unit 224, a second mark comparison unit 202, and a second mark comparison unit 220. A mark dictionary 204, an encoding unit 206, and a comparison parameter determination unit 226 are included. As described above, the image encoding program 22 has a configuration in which the first mark comparison unit 220, the first mark dictionary 222, and the deterioration determination unit 224 are added to the image encoding program 20 (FIG. 4). 9 that are substantially the same as those shown in FIG. 4 are denoted by the same reference numerals, and the second mark comparison unit 202 and the second mark dictionary 204 are the same as those shown in FIG. Corresponds to the mark comparison unit 202 and the mark dictionary 204 in the image encoding program 20.

  In the image encoding program 22, the first mark comparison unit 220 performs pattern matching processing on the marks cut out by the mark cutout unit 200. Specifically, the first mark comparison unit 220 determines whether or not the first mark dictionary 222 includes a mark that can be considered to match the cut mark, and the cut mark is included. In this case, the mark is replaced with a mark included in the first mark dictionary 222, and if not included, the mark is added to the first mark dictionary 222.

  Here, the first mark comparison unit 220 performs a pattern matching process using a comparison parameter having a smaller value than the comparison parameter used in the second mark comparison unit 202. Preferably, the first mark comparison unit 220 determines whether or not the extracted mark and the mark included in the first mark dictionary 222 to be compared completely match.

  The first mark comparison unit 220 outputs statistical information obtained by the first pattern matching process to the deterioration determination unit 224. The statistical information includes, for example, the number and size of the marks cut out from the input image, the frequency and the ratio at which it is determined that the cut out marks match the marks included in the first mark dictionary 222, and the like.

  In addition, the first mark comparison unit 220 may generate a hash value based on the cut mark and perform pattern matching processing based on the hash value. Specifically, the first mark comparison unit 220 determines whether or not the first mark dictionary 222 includes a hash value that matches the hash value generated based on the cut mark, and the hash If the value is included, the extracted mark is replaced with a mark corresponding to the hash value, and if not included, the extracted mark and the hash value are added to the first mark dictionary 222. For example, the first mark comparison unit 220 generates a hash value by applying a technique such as CRC (Cyclic Redundancy Check) or MD5 (Message Digest Algorithm 5).

  The mark dictionary 222 stores a plurality of different marks and indexes. Note that when the first mark comparison unit 220 generates a hash value, the mark dictionary 222 further stores the generated hash value.

  The deterioration determination unit 224 determines the deterioration state of the input image based on the statistical information output from the first mark comparison unit 220. Specifically, the degradation determination unit 224 determines whether the input image is an undegraded image or a degraded image. For example, when the deterioration determination unit 224 determines that the extracted mark matches the mark included in the first mark dictionary 222, the appearance frequency of the mark pair determined to match is equal to or greater than a predetermined threshold, and the like. The input image is determined to be a non-degraded image. Otherwise, the input image is determined to be a deteriorated image.

  In the document image, for example, a plurality of the same marks such as a word (or part thereof), a Japanese particle (or part thereof), a vowel (or part thereof) such as a and e of the English alphabet, , May be included. When the input image is a deteriorated image, even if the same mark repeatedly appears in the document, it is rare that these marks are determined to match completely as an image. Therefore, it is effective to determine the deterioration state of the input image based on the frequency with which it is determined that the mark cut out from the input image matches the stored mark.

  The deterioration determination unit 224 may determine the deterioration state of the input image based on statistical information regarding a mark cut out from the input image and including a number of pixels equal to or greater than a predetermined value. When the mark is configured with the number of pixels less than the predetermined value, the ratio that the mark is considered to match the mark stored in the first mark dictionary 222 is the ratio of the mark configured with the number of pixels that is greater than or equal to the predetermined value. It is because it may become larger.

  Therefore, the first mark comparison unit 220, the first mark dictionary 222, and the deterioration determination unit 224 constitute deterioration state acquisition means for acquiring the deterioration state of the binary input image based on the first pattern matching method. To do. The comparison parameter determination unit 226 determines a comparison parameter based on the deterioration state of the input image acquired in this way. The determination method is the same as that of the comparison parameter determination unit 208 in the first embodiment.

FIG. 10 is a diagram for explaining the difference between the calculation amount in the first pattern matching method by the first mark comparison unit 220 and the calculation amount in the second pattern matching method by the second mark comparison unit 202.
FIG. 10A illustrates a mark cut out from the input image. As shown in FIG. 10A, for example, the mark has a height of h pixels and a width of w pixels.

  FIG. 10B is a diagram schematically illustrating the search range of the second mark comparison unit 202. The second mark comparison unit 202 compares not only a mark having the same size as the cut mark but also a mark having a different size from the cut mark and the cut mark. For example, the cut mark is also compared with a mark that differs in at least one of height and width by one pixel. In this example, as shown in FIG. 10B, the comparison target marks range from a mark whose width and height are w-1 pixels and h-1 pixels to a mark that is w + 1 pixels and h + 1 pixels, respectively. Including. If N marks are stored for each size, the cut marks are compared N times for each size.

  FIG. 10C is a diagram schematically illustrating the search range of the first mark comparison unit 220. Since the first mark comparison unit 220 compares the cut out mark with a mark having the same size as the mark, the mark to be compared has a width and a height as shown in FIG. Only marks that are w pixels and h pixels, respectively. Since the calculation amount of the pattern matching process is O (N2), the calculation amount of the first pattern matching process is, for example, 1/81 of the calculation amount of the second pattern matching process.

FIG. 11 is a flowchart of the comparison parameter acquisition process (S30) by the image encoding program 22. 11 that are substantially the same as those shown in FIG. 6 are denoted by the same reference numerals.
As shown in FIG. 11, in the processes of S104 and S106, a binary input image is received by the image encoding program 22, and a mark is cut out from the input image.

In step 200 (S200), the first mark comparison unit 220 generates a hash value based on the mark cut out by the mark cutout unit 200.
In step 202 (S202), the first mark comparison unit 220 performs comparison processing based on the generated hash value.

  In step 204 (S204), the first mark comparison unit 220 determines whether or not the first mark dictionary 222 includes a hash value that matches the generated hash value. The first mark comparison unit 220 proceeds to the process of S206 if a matching hash value is included, and proceeds to the process of S208 otherwise.

In step 206 (S206), the first mark comparison unit 220 stores the clipped mark in the first mark dictionary 222 as a mark corresponding to the hash value generated based on the mark. Replace with mark.
In step 206 (S206), the first mark comparison unit 220 adds the extracted mark and hash value to the first mark dictionary 222, and adds an index to the mark.

  In step 210 (S210), the mark cutout unit 200 determines whether or not all the marks included in the input image have been cut out. The mark cutout unit 200 proceeds to the process of S212 when all the marks have been cut out, and returns to the process of S106 otherwise.

In step 212 (S212), the first mark comparison unit 220 generates statistical information related to the first pattern matching process.
In step 214 (S214), the deterioration determination unit 224 determines the deterioration state of the input image based on the generated statistical information.
In step 216 (S216), a comparison parameter is determined based on the determination result by the deterioration determination unit 224.

  Thereafter, the parameters acquired in this way are used, and the second pattern matching process and the encoding process are executed in the same manner as in the first embodiment. Note that the first mark comparison unit 220 may perform the pattern matching process by comparing the extracted mark and the stored mark without generating a hash value.

Next, an image encoding device 2 according to the third embodiment of the present invention will be described.
The image encoding device 2 according to the present embodiment relates to the second embodiment in that the second pattern matching process is executed for a mark that is determined not to match in the first pattern matching process. This is different from the image encoding device 2.

FIG. 12 is a diagram illustrating a functional configuration of the image encoding program 24 executed by the control device 100 of the image encoding device 2 according to the present embodiment. Of the components shown in FIG. 12, those substantially the same as those shown in FIG. 9 are denoted by the same reference numerals.
As shown in FIG. 12, the image encoding program 24 includes a first mark comparison unit 220 and a second mark comparison unit 202 of the image encoding program 22 (FIG. 9), respectively. The second mark comparison unit 242 is replaced.

  In the image encoding program 24, the first mark comparison unit 240 performs the processing of the first mark comparison unit 220 in the image encoding program 22, and sets the mark index and drawing position to the second mark comparison unit 242. Output.

  The second mark comparison unit 242 reads a mark from the first mark dictionary 222, receives a mark index and a drawing position from the first mark comparison unit 240, and performs a second pattern matching process on the read mark. Go to generate the second mark dictionary 204. Therefore, the number of marks that are targets of the second pattern matching process is smaller than the number of marks that are targets of the first pattern matching process.

  For example, the mark cutout unit 200 cuts out six marks “A”, “A”, “A”, “B”, “B”, and “C” from the input image, and the first mark comparison unit 240 If it is determined that three “A” s match and two “B” s also match, the first mark dictionary 222 stores three marks “A”, “B”, and “C”. Yes. In this case, the second mark comparison unit 242 reads only these three marks from the first mark dictionary 222 and performs a second pattern matching process on the three marks.

Next, an image encoding device 2 according to the fourth embodiment of the present invention will be described.
When it is determined that the input image is an undegraded image, the image encoding device 2 according to the present embodiment does not execute the second pattern matching process, and encodes the result of the first pattern matching process. This is different from the image encoding device 2 according to the third embodiment.

FIG. 13 is a diagram illustrating a functional configuration of the image encoding program 26 executed by the control device 100 of the image encoding device 2 according to the present embodiment. Of the components shown in FIG. 13, components substantially the same as those shown in FIG. 12 are denoted by the same reference numerals.
As shown in FIG. 13, in the image encoding program 26, the comparison parameter determination unit 226 is deleted from the image encoding program 24 (FIG. 12), and the first selection unit 260 and the second selection unit 262 are added. It has a configuration.

  In the image encoding program 26, the first selection unit 260 receives the determination result by the deterioration determination unit 224, and changes the output destination based on the determination result. Specifically, when it is determined that the input image is a non-degraded image, the first selection unit 260 outputs the mark read from the first mark dictionary 222 to the encoding unit 206, and the input image Is determined to be a deteriorated image, the mark is output to the second mark comparison unit 242.

  The second selection unit 262 also receives the determination result by the deterioration determination unit 224 and changes the output destination based on the determination result. Specifically, when it is determined that the input image is a non-degraded image, the second selection unit 262 sends the mark index and the drawing position output from the first mark comparison unit 240 to the encoding unit 206. In contrast, when it is determined that the input image is a deteriorated image, the information is output to the second mark comparison unit 242.

  Therefore, when the input image is an undegraded image, the second pattern matching process is skipped. The second mark comparison unit 242 performs a second pattern matching process using a preset comparison parameter.

Next, an image encoding device 2 according to the fifth embodiment of the present invention will be described.
The image encoding device 2 according to the present embodiment is different from the image encoding device 2 according to the fourth embodiment in that a deterioration determination process is performed based on the presence / absence of information regarding deterioration of an input image.

FIG. 14 is a diagram illustrating a functional configuration of the image encoding program 28 executed by the control device 100 of the image encoding device 2 according to the present embodiment. 14 that are substantially the same as those shown in FIG. 13 are denoted by the same reference numerals.
As shown in FIG. 14, the image encoding program 28 has a configuration in which the deterioration determining unit 224 of the image encoding program 26 (FIG. 13) is replaced with a deterioration determining unit 280.

  In the image encoding program 28, the degradation determination unit 280 receives information related to degradation of the input image in the same manner as the comparison parameter determination unit 208 in the first embodiment. This information includes whether the input image is a degraded image, an undegraded image, or the degradation state is unknown. When the deterioration determination unit 280 receives information regarding the deterioration of the input image, acquires the deterioration state of the input image from the information, and does not receive the information or receives information indicating that the deterioration state is unknown The presence or absence of deterioration is determined in the same manner as the deterioration determination unit 224 in the second to fourth embodiments. The degradation determination unit 224 outputs the determination result to the first selection unit 260 and the second selection unit 262.

  For example, when the image encoding device 2 according to the present embodiment is included in a multifunction device in which a scanner and a printer are integrated, the image encoding device 2 treats a read image output from the scanner as a deteriorated image, A print image transmitted from a computer or the like (not shown) is treated as an undegraded image. When there is a possibility that the read image read by the scanner or the like is pasted on such a printed image, the image encoding device 2 is assumed to have an unknown deterioration state.

FIGS. 1A to 1C exemplify a deteriorated image and a non-degraded image that have undergone pattern matching processing, and FIGS. F) illustrates the case of an undegraded image. It is a figure which shows the input image for demonstrating degradation of the image quality in a degradation image and a non-degradation image. It is a figure which shows the hardware constitutions of the image coding apparatus 2 and the image decoding apparatus 4 which concern on embodiment of this invention centering on the control apparatus 100. FIG. It is a figure which shows the function structure of the image coding program 20 performed by the control apparatus 100 of the image coding apparatus 2. FIG. FIG. 5A is a diagram illustrating a relationship between an input image, a mark dictionary 204, and an encoding process. FIG. 5A illustrates a binary input image, and FIG. 5B is generated based on the input image. The mark dictionary 204 is illustrated, and FIG. 5C illustrates information regarding the mark drawing position. It is a figure which shows the flowchart of the image coding process (S10) performed by the image coding program 20. FIG. It is a figure which shows the function structure of the image decoding program 40 performed by the control apparatus 100 of the image decoding apparatus 4. FIG. It is a figure which shows the flowchart of the image decoding process (S20) by the image decoding program 40. FIG. It is a figure which shows the function structure of the image coding program 22 performed by the control apparatus 100 of the image coding apparatus 2 which concerns on this embodiment. It is a figure explaining the difference in the calculation amount in the 1st pattern matching method by the 1st mark comparison part 220, and the calculation amount in the 2nd pattern matching method by the 2nd mark comparison part 202. FIG. It is a figure which shows the flowchart of the comparison parameter acquisition process (S30) by the image coding program 22. FIG. It is a figure which shows the function structure of the image coding program 24 performed by the control apparatus 100 of the image coding apparatus 2 which concerns on this embodiment. It is a figure which shows the function structure of the image coding program 26 performed by the control apparatus 100 of the image coding apparatus 2 which concerns on this embodiment. It is a figure which shows the function structure of the image coding program 28 performed by the control apparatus 100 of the image coding apparatus 2 which concerns on this embodiment.

Explanation of symbols

2 Image Encoding Device 4 Image Decoding Device 20 Image Encoding Program 22 Image Encoding Program 24 Image Encoding Program 26 Image Encoding Program 28 Image Encoding Program 40 Image Decoding Program 100 Control Device 200 Mark Extraction Unit 202 Mark Comparison Unit 204 mark dictionary 206 encoding unit 208 comparison parameter determination unit 220 mark comparison unit 222 mark dictionary 224 deterioration determination unit 226 comparison parameter determination unit 240 mark comparison unit 242 mark comparison unit 260 first selection unit 262 second selection unit 280 Degradation determining unit 400 Code data receiving unit 402 Decoding processing unit 404 Decoded image generating unit

Claims (17)

Degradation state accepting means for accepting an input relating to the degradation state of the image;
A dictionary generation means for generating an image pattern dictionary from an image pattern included in a binary input image based on a pattern matching method corresponding to the deterioration state received by the deterioration state reception means;
An image encoding device comprising: an image pattern dictionary generated by the dictionary generation means; and an encoding means for encoding a position at which an image pattern included in the image pattern dictionary appears in an input image. A deterioration state acquisition means for acquiring a deterioration state of a binary input image based on a first pattern matching method;
A dictionary generation unit that generates an image pattern dictionary from an image pattern included in an input image based on a second pattern matching method corresponding to the deterioration state acquired by the deterioration state acquisition unit;
An image encoding device comprising: an image pattern dictionary generated by the dictionary generation means; and an encoding means for encoding a position at which an image pattern included in the image pattern dictionary appears in an input image. The image coding apparatus according to claim 2, wherein the deterioration state acquisition unit determines whether or not an image pattern included in an input image matches an image pattern to be compared as the first pattern matching. The image coding apparatus according to claim 2, wherein the deterioration state acquisition unit performs pattern matching based on a hash value generated from an image pattern. 5. The deterioration state acquisition unit determines a deterioration state of an input image based on a result of pattern matching performed on an image pattern including a number of pixels equal to or greater than a predetermined value. Image coding apparatus. The image encoding according to any one of claims 2 to 5, wherein when the input image is a non-degraded image, the dictionary generating means applies pattern matching by applying a comparison parameter smaller than that when the input image is a deteriorated image. apparatus. The deterioration state acquisition means generates a first image pattern dictionary from an image pattern included in a binary input image,
The image encoding according to any one of claims 2 to 5, wherein the dictionary generation unit generates a second image pattern dictionary from an image pattern included in the first image pattern dictionary generated by the deterioration state acquisition unit. apparatus. When the input image is a non-degraded image, the encoding unit generates a first image pattern dictionary generated by the deterioration state acquiring unit and an image pattern included in the first image pattern dictionary appear in the input image. And when the input image is a degraded image, the second image pattern dictionary generated by the dictionary generation means, and the position at which the image pattern included in the second image pattern dictionary appears in the input image, The image encoding device according to claim 7. A deterioration state receiving means for receiving an input regarding the deterioration state of the image;
When the deterioration state is received by the deterioration state reception unit, the encoding unit includes a first image pattern dictionary generated by the deterioration state acquisition unit and an image pattern included in the first image pattern dictionary. The image encoding apparatus according to claim 8, wherein a position appearing in the input image is encoded. An image pattern dictionary is obtained from image patterns included in a binary input image based on a degradation state acceptance unit that accepts an input related to a degradation state of an image, and a pattern matching method corresponding to the degradation state accepted by the degradation state acceptance unit. An image encoding device comprising: a dictionary generating means for generating; an image pattern dictionary generated by the dictionary generating means; and an encoding means for encoding a position at which an image pattern included in the image pattern dictionary appears in an input image. Code data receiving means for receiving code data encoded by
Decoding means for decoding the code data received by the code data receiving means;
An image decoding apparatus comprising: a decoded image generation unit configured to generate a decoded image based on the data decoded by the decoding unit. Based on a first pattern matching method, a deterioration state acquisition unit that acquires a deterioration state of a binary input image, and a second pattern matching method that corresponds to the deterioration state acquired by the deterioration state acquisition unit. A dictionary generation unit that generates an image pattern dictionary from an image pattern included in the input image, an image pattern dictionary generated by the dictionary generation unit, and a position at which the image pattern included in the image pattern dictionary appears in the input image. Code data receiving means for receiving code data encoded by an image encoding device having encoding means for encoding;
Decoding means for decoding the code data received by the code data receiving means;
An image decoding apparatus comprising: a decoded image generation unit configured to generate a decoded image based on the data decoded by the decoding unit. Degradation state accepting means for accepting an input relating to the degradation state of the image;
A dictionary generation means for generating an image pattern dictionary from an image pattern included in a binary input image based on a pattern matching method corresponding to the deterioration state received by the deterioration state reception means;
Encoding means for encoding the image pattern dictionary generated by the dictionary generation means and the position where the image pattern included in the image pattern dictionary appears in the input image;
Decoding means for decoding the code data encoded by the encoding means;
An image processing system comprising: decoded image generation means for generating a decoded image based on the data decoded by the decoding means. A deterioration state acquisition means for acquiring a deterioration state of a binary input image based on a first pattern matching method;
A dictionary generation unit that generates an image pattern dictionary from an image pattern included in an input image based on a second pattern matching method corresponding to the deterioration state acquired by the deterioration state acquisition unit;
Encoding means for encoding the image pattern dictionary generated by the dictionary generation means and the position where the image pattern included in the image pattern dictionary appears in the input image;
Decoding means for decoding the code data encoded by the encoding means;
An image processing system comprising: decoded image generation means for generating a decoded image based on the data decoded by the decoding means. In an image encoding device including a computer,
A degradation state accepting step for accepting input relating to the degradation state of the image;
A dictionary generation step of generating an image pattern dictionary from an image pattern included in a binary input image based on a pattern matching method corresponding to the accepted deterioration state;
A program that causes the image encoding device to execute the encoding step of encoding the generated image pattern dictionary and a position at which an image pattern included in the image pattern dictionary appears in an input image. In an image encoding device including a computer,
A deterioration state acquisition step of acquiring a deterioration state of a binary input image based on a first pattern matching method;
A dictionary generation step of generating an image pattern dictionary from an image pattern included in the input image based on the second pattern matching method corresponding to the acquired deterioration state;
A program that causes the image encoding device to execute the encoding step of encoding the generated image pattern dictionary and a position at which an image pattern included in the image pattern dictionary appears in an input image. In an image decoding apparatus including a computer,
In an image encoding apparatus including a computer, based on an image pattern included in a binary input image, based on a deterioration state receiving step for receiving an input regarding an image deterioration state and a pattern matching method corresponding to the received deterioration state Encoding is performed by performing a dictionary generation step for generating an image pattern dictionary, and an encoding step for encoding the generated image pattern dictionary and a position at which an image pattern included in the image pattern dictionary appears in an input image. A code data receiving step for receiving the encoded data;
A decoding step of decoding the received code data;
A program that causes a computer of the image decoding apparatus to execute a decoded image generation step of generating a decoded image based on the decoded data. In an image decoding apparatus including a computer,
In an image encoding device including a computer, based on a first pattern matching method, a deterioration state acquisition Sutepu that acquires a deterioration state of a binary input image, and a second pattern matching corresponding to the acquired deterioration state Based on the technique, a dictionary generation step of generating an image pattern dictionary from an image pattern included in the input image, the generated image pattern dictionary, and a position where the image pattern included in the image pattern dictionary appears in the input image. A code data receiving step for receiving code data encoded by performing an encoding step for encoding;
A decoding step of decoding the received code data;
A program that causes a computer of the image decoding apparatus to execute a decoded image generation step of generating a decoded image based on the decoded data.

Images