JP2007087190A - Image processing device and program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To carry out image processing using wavelet transformation at high speed. <P>SOLUTION: By means of this image processing device, it is determined whether attributes of respective image blocks are a background or not based on a wavelet coefficient of an LL component obtained by performing wavelet transformation on image data. As to the image block determined as the background, the wavelet coefficients of HL, LH, and HH components are replaced by predetermined values. As to the image block not determined as the background, an edge is detected by referring to the HL, LH, and HH components and edge processing is carried out by multiplying the wavelet coefficient corresponding to the edge by a predetermined edge enhancement coefficient, and then, reverse wavelet transformation is carried out for restoring the image data and output them. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to a technique for performing image processing on a scanned image and outputting the image, and more particularly to a technique for performing image processing according to the characteristics of each object constituting the scanned image and outputting the object.

Generally, an image read by a scanner device or a copying machine (hereinafter referred to as a scanned image) is subjected to image processing such as color conversion processing, and the resulting image is formed on a recording material such as printing paper and output. ing. Some scanned images have image areas having different attributes such as characters, patterns, backgrounds (hereinafter, characters, patterns, backgrounds, etc. are also referred to as “attributes”). When image processing is performed, it is desirable to perform image processing suitable for the attribute of each image region. For example, for an area whose attribute is a character, it is desirable to perform image processing that emphasizes the edge of the character in order to avoid blurring the outline of the character (hereinafter also referred to as an edge).
As an example of a technique that enables this, there are techniques disclosed in Patent Documents 1 to 4. These patent documents disclose techniques for performing image quality enhancement processing such as edge enhancement on an image to be processed using wavelet transform.

FIG. 15 is a diagram for explaining the wavelet transform and the inverse wavelet transform.
As shown in FIG. 15, in the wavelet transform, first, “H ↓ 2 filter” and “G ↓ 2 filter” shown in FIG. 15 are predetermined for image data in the order of the main scanning direction and the sub-scanning direction. Further, wavelet coefficients representing the LL component, the LH component, the HL component, and the HH component of the original image represented by the image data are calculated by performing filter processing in combination with the four combination patterns (see FIG. 15). On the other hand, in the inverse wavelet transform, each wavelet coefficient is input, and in the order of the sub-scanning direction and the main scanning direction, “↑ 2Q filter” and “↑ 2P filter” shown in FIG. The image data is restored by performing the filtering process in combination in FIG. Each wavelet coefficient generated by the wavelet transform has the following characteristics. In the HH component, an edge component in the lick direction (45 ° and 135 ° directions) appears strongly. In the LH component, a lateral edge component appears strongly. In the HL component, a vertical edge component appears strongly. The LL component has substantially the same characteristics as an image obtained by reducing the current image by 1/2 in the vertical and horizontal directions. “H ↓ 2 filter”, “G ↓ 2 filter”, “↑ 2Q filter” and “↑ 2P filter” are filters called Haar filters.
JP-A-6-274614 JP-A-6-223172 JP 2002-247368 A JP 2003-9153 A

By the way, in the technique disclosed in each of the above patent documents, the edge detection is performed with reference to all other wavelet coefficients excluding the LL component at the time of edge detection prior to edge enhancement, so that the processing speed is increased. There is a problem that it is difficult.
The present invention has been made in view of the above problems, and an object of the present invention is to provide a technique that enables high-speed image processing using wavelet transform.

In order to solve the above-mentioned problems, the present invention provides an input means for inputting image data representing an image with lightness and saturation of each pixel, wavelet transform is performed on the image data, and an image represented by the image data is displayed. Wavelet transform means for generating wavelet coefficients of LL component, LH component, HL component and HH component for each of the lightness component and the saturation component, and an image represented by the wavelet coefficients of the LL component for the lightness component in a plurality of blocks For each of the blocks, a determination unit that determines whether the attribute is a background by referring to a wavelet coefficient of a corresponding LL component, and a block that is determined to be a background by the determination unit Write each wavelet coefficient of the HL component, LH component and HH component corresponding to the block to a predetermined value On the other hand, for a block that is determined not to be a background by the determination means, an edge portion is detected by referring to the wavelet coefficients of the HL component, LH component, and HH component corresponding to the block, and a predetermined edge enhancement coefficient The inverse wavelet transform is applied to the edge processing means for multiplying the LL component, the wavelet coefficient of the LL component, and the wavelet coefficients of the HL component, the LH component, and the HH component that have been processed according to the attribute by the edge processing means. There is provided an image processing apparatus characterized by comprising output means for restoring image data and outputting the image data.
According to such an image processing device, for a block determined to be a background by referring to the wavelet coefficient of the LL component, without referring to the wavelet coefficients of the LH, HL, and HH components of the block, For those blocks whose wavelet coefficients are rewritten to predetermined values and are determined to be other than the background, the edge portion is detected by referring to the wavelet coefficients of the HL component, LH component and HH component corresponding to the block. These coefficients corresponding to the edge portion are multiplied by a predetermined edge enhancement coefficient.

In a more preferred aspect, the determining means determines whether the attribute of the block determined not to be a background is a character, an image, or unknown, and the edge processing means For the block whose attribute is determined to be a character by the determination means, the wavelet coefficient of the HL component, the LH component and the HH component corresponding to the edge portion is multiplied by a first edge enhancement coefficient, and the attribute is a character. For the block determined by the determining means, the second edge enhancement coefficient whose value is smaller than the first edge enhancement coefficient in the wavelet coefficients of the HL component, LH component and HH component corresponding to the edge portion. For the block whose attribute is determined to be unknown, the block corresponding to the edge portion Each of the wavelet coefficients of the HL component, the LH component, and the HH component is compared with a predetermined threshold value. When the threshold value is larger than the threshold value, the first edge enhancement coefficient is multiplied. The second emphasis coefficient is multiplied.
In such an aspect, for a block in which the attribute is determined to be a character or a block in which the attribute is determined to be a pattern, the first or second edge enhancement coefficient is set in the block according to the attribute. For a block that has been multiplied by the wavelet coefficients of the HL component, LH component, and HH component corresponding to the edge portion and whose attribute is determined to be unknown, the HL component, LH component, and HH corresponding to the edge portion in the block If each wavelet coefficient of the component has a value greater than a predetermined threshold (that is, when the edge strength is strong), the first edge enhancement coefficient (that is, the edge enhancement coefficient for characters) is multiplied, and conversely When the value is smaller than a predetermined threshold (that is, when the edge strength is weak), the second edge enhancement coefficient (that is, the pattern edge enhancement coefficient) It is multiplied.

In order to solve the above problems, the present invention receives image data representing an image with brightness and saturation of each pixel, and applies wavelet transform to the image data to represent the image data. A first step of generating wavelet coefficients of the LL component, LH component, HL component, and HH component for each of the lightness component and the saturation component of the image, and a plurality of images represented by the wavelet coefficients of the LL component for the lightness component A second step of determining whether the attribute is a background by referring to the wavelet coefficient of the corresponding LL component for each of the blocks, and the second step if the attribute is a background For the determined block, the wavelet coefficients of the HL component, the LH component, and the HH component corresponding to the block. For the block determined in the second step that is not the background, the edge portion is referenced with reference to the wavelet coefficients of the HL component, LH component, and HH component corresponding to the block. And a predetermined edge enhancement coefficient is multiplied by a third step, the wavelet coefficient of the LL component, and the HL component, the LH component, and the HH component subjected to processing according to the attribute in the third step And a fourth step of restoring the image data by applying inverse wavelet transform to the wavelet coefficients of the image and outputting the image data.
According to such a program, by installing this program in a general computer apparatus, it is possible to give the same function as the image processing apparatus to the computer apparatus.

  In order to solve the above problems, the present invention provides an input means for inputting image data representing an image with brightness and saturation of each pixel, and wavelet transform is applied to the image data input to the input means. , Wavelet transform means for generating wavelet coefficients of LL component, LH component, HL component and HH component for each of the lightness component and saturation component of the image represented by the image data, and an edge referring to the wavelet coefficient of the lightness component Brightness component processing means for detecting the intensity and direction of the image, and multiplying the wavelet coefficients of the LH component, HL component, and HH component corresponding to the edge by an edge enhancement coefficient corresponding to the intensity, and the wavelet coefficient of the saturation component Of the LL component, the saturation calculated from the LL component and the lightness component processing means The wavelet coefficient corresponding to the detected edge is multiplied by a predetermined value in accordance with the detected intensity of the edge and subjected to saturation conversion, while the LH component, HL component, and HH component are detected by the lightness component processing means. A saturation component processing means for rewriting a wavelet coefficient value corresponding to a region determined according to the determined edge direction to a predetermined value, a processing result by the lightness component processing means, and a processing result by the saturation component processing means in an inverse wavelet There is provided an image processing apparatus characterized by having an output means for performing conversion to restore image data and outputting the image data.

  In order to solve the above problems, the present invention receives image data representing an image with brightness and saturation of each pixel, and applies wavelet transform to the image data to represent the image data. A first step of generating wavelet coefficients of LL component, LH component, HL component and HH component for each of the lightness component and the saturation component of the image, and the intensity and direction of the edge with reference to the wavelet coefficient of the lightness component A second step of detecting and multiplying the LH component corresponding to the edge, the wavelet coefficient of the HL component and the HH component by an edge enhancement coefficient corresponding to the intensity, and the LL component of the wavelet coefficients of the saturation component Is the saturation calculated from the LL component and the intensity of the edge detected in the second step The wavelet coefficient corresponding to the edge is multiplied by a predetermined value in response to the saturation conversion, and the LH component, the HL component, and the HH component are determined according to the edge direction detected in the second step. A third step of rewriting a wavelet coefficient value corresponding to a predetermined region to a predetermined value, and performing inverse wavelet transform on the processing result of the second step and the processing result of the third step to restore image data And a fourth step of outputting the image data is provided.

  According to the present invention, it is possible to perform image processing using wavelet transformation at high speed.

The best mode for carrying out the present invention will be described below with reference to the drawings.
(A. First embodiment)
(A-1. Configuration)
FIG. 1 is a block diagram illustrating a configuration example of an image processing system including an image processing apparatus 200 according to an embodiment of the present invention. As shown in FIG. 1, the image processing system includes an image reading apparatus 100, an image processing apparatus 200, and an image forming apparatus 300. The image processing apparatus 200 is, for example, a USB (Universal Serial Bus). Communication lines 400 a and 400 b such as cables are connected to the image reading apparatus 100 and the image forming apparatus 300.

  The image reading apparatus 100 is a scanner apparatus having an automatic paper feed mechanism such as an ADF (Auto Document Feeder), for example, reads a paper document set in the ADF page by page, and obtains RGB image data corresponding to the read image. The image is transferred to the image processing apparatus 200 via the communication line 400a. Here, the RGB image data is image data representing an image in which colors are represented by a combination of the three primary colors of light, red (R), green (G), and blue (B).

  The image processing device 200 performs edge enhancement processing for converting the RGB image data delivered via the communication line 400 to L * a * b * image data to enhance edges, and then converts the image data into YMCK image data for communication. The image is transferred to the image forming apparatus 300 via the line 400b. Here, L * a * b * image data is image data in which the color of each pixel is represented by lightness (L) and saturation (a, b), and YMCK image data is the color of each pixel. Is image data representing an image in which colors are represented by combinations of four colors of yellow (Y), magenta (M), cyan (C), and black (K), which are basic colors for printing. The configuration and operation of the image processing apparatus 200 will be described in detail later.

  The image forming apparatus 300 is, for example, an electrophotographic printer apparatus, and an image corresponding to the YMCK image data delivered from the image processing apparatus 200 via the communication line 400b is electrophotographic to a recording material such as printing paper. Form and output. In the present embodiment, the case where the image forming apparatus 300 is an electrophotographic printer apparatus will be described. However, a printer apparatus that forms an image using another system such as an ink jet system may be used.

  In this embodiment, a case where the image processing apparatus 200 is connected to the image reading apparatus 100 or the image forming apparatus 300 via a USB cable will be described. However, a LAN (Local Area Network) or a WAN (Wide Area Network) is described. Of course, the image processing apparatus 200 may be connected to the image reading apparatus 100 or the image forming apparatus 300 via a communication network such as the Internet. In this embodiment, the case where the image reading apparatus 100, the image processing apparatus 200, and the image forming apparatus 300 are configured by individual hardware will be described. However, these may be configured by integral hardware. Of course. In such an aspect, an internal bus connecting the image reading apparatus 100, the image processing apparatus 200, and the image forming apparatus 300 in the hardware plays the role of the communication line.

FIG. 2 is a diagram illustrating an example of a hardware configuration of the image processing apparatus 200.
As shown in FIG. 2, the image processing apparatus 200 includes a control unit 210, an interface (hereinafter, IF) unit 220, a storage unit 230, and a bus 240 that mediates data exchange between these components. .

  The control unit 210 is, for example, a CPU (Central Processing Unit), and centrally controls each unit of the image processing apparatus 200 by executing a program stored in the storage unit 230.

  The IF unit 220 is a USB interface, to which communication lines 400a and 400b are connected. The IF unit 220 receives the RGB image data sent from the image reading apparatus 100 via the communication line 400a and delivers it to the control unit 210, while the YMCK image data delivered from the control unit 210 is sent via the communication line 400b. To the image forming apparatus 300. When the image reading apparatus 100 or the image forming apparatus 300 is connected via a communication network such as a LAN, the IF unit 220 may be configured by a NIC (Network Interface Card).

As shown in FIG. 2, the storage unit 230 includes a nonvolatile storage unit 230a and a volatile storage unit 230b.
The nonvolatile storage unit 230a is, for example, a hard disk, and the volatile storage unit 230b is, for example, a RAM (Random Access Memory). The volatile storage unit 230b is used as a work area by the control unit 210 operating according to a program. On the other hand, the non-volatile storage unit 230a stores in advance a program for causing the control unit 210 to perform image processing characteristic of the image processing apparatus according to the present invention. In addition, the nonvolatile storage unit 230a stores color conversion processing from RGB image data to L * a * b * image data (hereinafter, pre-stage color conversion processing) or L * a * b * image data to YMCK image data. A look-up table (hereinafter referred to as LUT) referred to during the execution of the color conversion process (hereinafter referred to as the subsequent color conversion process) and the image represented by the input image data are divided into a plurality of blocks, and the attributes of these blocks are characters. , An attribute discrimination table to be referred to when discriminating between the pattern and the background is stored in advance.

  In the attribute discrimination table, data representing a reference pattern of a lightness (L) histogram and a saturation (a, b) histogram according to the attributes of image materials (characters, pictures, etc.) such as characters and photographs (or pictures). Is stored for each attribute. For example, FIG. 3A is a black character, FIG. 3B is a color character, FIG. 3C is a color photograph (picture), FIG. 3D is a black-and-white photograph (picture), and FIG. The above-mentioned reference pattern.

  As shown in FIG. 3A, the reference pattern for black characters is a lightness histogram with a highlight and a peak at high density and a saturation histogram with one peak at the center. As shown in FIG. 3 (b), there are a brightness histogram having peaks in highlight and high density, and a saturation histogram having two peaks at positions shifted from the center and the center. As shown in FIG. 3C, the reference pattern for the color pattern is a lightness histogram with a gentle peak and an indefinite peak, and the saturation histogram with a gentle peak indefinite, and the reference pattern for a monochrome pattern is shown in FIG. As shown in d), it is a lightness histogram with gentle peaks and an indefinite peak and a saturation histogram with one peak in the center. As shown in FIG. 3E, the reference pattern for the background is a lightness histogram having one peak in the highlight and a saturation histogram having one peak in the center.

  As described above, since the data representing the reference pattern for each attribute is stored for each attribute in the attribute determination table, the control unit 210 calculates the reference pattern for each object, and the calculation result is the above-described result. By comparing with the contents stored in the attribute determination table, the attribute of each object can be determined. In the present embodiment, the case where the lightness histogram and the saturation histogram are used as the reference pattern for each attribute has been described. However, for example, the lightness histogram and the saturation histogram are used as the reference pattern such as the on / off inversion frequency of the pixel value. Is of course good.

  The above is the configuration of the image processing apparatus 200. As described above, the hardware configuration of the image processing apparatus 200 is the same as the hardware configuration of a general computer apparatus, and the control unit 210 is operated according to the program stored in the nonvolatile storage unit 230a. Functions specific to the image processing apparatus according to the above are realized.

(A-2. Operation)
Next, edge enhancement processing executed by the control unit 210 operating according to the program will be described with reference to the drawings. Note that the pre-stage color conversion process and the post-stage color conversion process executed by the control unit 210 are not different from those performed by the conventional image processing apparatus. Therefore, the following is characteristic of the image processing apparatus according to the present invention. The edge enhancement process will be described.
FIG. 4 is a flowchart showing the flow of edge enhancement processing executed by the control unit 210 in accordance with the program. As shown in FIG. 4, the control unit 210 first uses the Haar filter described above for the L * a * b * image data generated in the previous color conversion process (in detail, discrete wavelet conversion). And wavelet coefficients representing the LL, LH, HL, and HH components of the image represented by the image data are calculated (step SA100).

Next, the control unit 210 refers to the wavelet coefficient of the LL component among the wavelet coefficients generated in step SA100, and the image represented by the wavelet coefficient of the LL component is an image block having a predetermined size (for example, 1 × 1 mm). (Step SA110). Next, the control unit 210 calculates a lightness (L) histogram and a saturation (a, b) histogram for each image block divided in step SA110, and stores them in the attribute determination table described above. By comparing with the reference pattern, the attribute of each image block is determined (step SA120). In this case, for example, the control unit 210 calculates the degree of deviation from the reference pattern by an appropriate calculation method, and selects the reference pattern having the smallest degree of deviation. Note that the control unit 210 determines that the attribute of the image block is “unknown” when there is no large difference in the degree of deviation from each reference pattern.
In the present embodiment, the case where the size of the image block is set to 1 × 1 mm has been described. However, it is possible to more accurately determine the attribute as the size is reduced, but it takes time to determine the attribute. End up. Therefore, it is preferable that the required processing speed and image quality are comprehensively determined and the size of the image block is set appropriately.

Next, the control unit 210 performs edge processing according to the attribute determined in step SA120 on each image block divided in step SA110 (step SA130).
For example, for an image block whose attribute is determined to be a background (hereinafter referred to as a background block), the control unit 210 has all the LH, HL, and HH wavelet coefficients corresponding to the image block without edges, that is, blank or paper white. Is rewritten to a predetermined value (eg, “0”). Since the edge component appearing in the background portion is generally noise generated in the image reading process by a scanner device or the like, the LH, HL, and HH wavelet coefficients are all rewritten to predetermined values, Such noise is expected to be removed.

  For an image block whose attribute is determined to be a character or a picture, the control unit 210 performs edge processing shown in FIG. As shown in FIG. 5, the control unit 210 first identifies regions in the LH, HL, and HH components corresponding to the image block (step SB100), and performs edge detection in those regions (step SB110). . More specifically, the control unit 210 compares the value of the wavelet coefficient of each of the LH, HL, and HH components corresponding to the region with a predetermined edge determination threshold value. The corresponding wavelet coefficient is detected.

  Next, the control unit 210 multiplies the wavelet coefficient detected in step SB110 by an edge enhancement coefficient corresponding to the attribute of the image block (step SB120). Specifically, the control unit 210 multiplies a character edge enhancement coefficient for a wavelet coefficient corresponding to an image block (hereinafter referred to as a character block) whose attribute is determined to be a character, and the attribute is a pattern. A wavelet coefficient corresponding to the discriminated image block (hereinafter referred to as a “pattern block”) is multiplied by a pattern edge enhancement coefficient. In the present embodiment, the edge enhancement coefficient for characters is set to a value larger than the edge enhancement coefficient for patterns. Therefore, for character blocks, the edges are emphasized so that the characters stand out. The edge is processed so as not to stand out.

  For an image block whose attribute is determined to be unknown (hereinafter, unknown block), the control unit 210 performs edge processing shown in FIG. As shown in FIG. 6, in the edge processing for the unknown block, the regions in the LH, HL and HH components corresponding to the unknown block are specified (step SB100), and the edge detection in those regions (step SB110). ) Is the same as the edge processing for the character or design block described above, except that the processing after step SB130 is executed after step SB110.

In FIG. 6, in step SB130 executed subsequent to step SB110, the control unit 210 determines whether or not the edge strength detected in step SB110 exceeds a predetermined threshold value. Then, when the determination result in step SB130 is “Yes”, the control unit 210 assumes that the character block is equivalent to the edge and multiplies the wavelet coefficient detected in step SB110 by the character edge enhancement coefficient. (Step SB140). On the other hand, if the determination result in step SB130 is “No”, the wavelet coefficient detected in step SB110 is multiplied by the design edge enhancement coefficient, assuming that it is equivalent to the edge for the design block (step SB150). ).
The above is the edge processing applied to each image block.

Returning to FIG. 4, when the edge processing in step SA <b> 130 is completed, the control unit 210 performs edge processing according to the wavelet coefficient of the LL component generated in step SA <b> 100 and the attribute of the image block in step SA <b> 130. The wavelet coefficients of the LH component, the HL component, and the HH component are input, and the inverse wavelet transform is executed using the Haar filter described above to restore the L * a * b * image data (step SA140), and this edge enhancement processing is completed. To do.
Thereafter, the control unit 210 converts the L * a * b * image data obtained by the edge enhancement processing described above to YMCK image data by performing the above-described post-color conversion processing, and forms the YMCK image data as an image forming unit. The image is transferred to the apparatus 300 and an image corresponding to the YMCK image data is output.

(A-3: Effect of the first embodiment)
As described above, according to the image processing apparatus 200 according to the first embodiment of the present invention, only the coefficients of the LL component among the coefficients of the LL, LH, HL, and HH components obtained by the wavelet transform are obtained. The background block and the blocks other than the background (that is, characters, pictures, and unknown blocks) are discriminated, and the former is not subjected to edge detection, so it is faster than the conventional technique. It becomes possible.
As for edge enhancement, the image processing apparatus 200 according to the first embodiment of the present invention does not use a digital filter such as an edge enhancement filter, and performs coefficient calculation (that is, sets wavelet coefficients corresponding to edges to image block attributes). Therefore, the speed can be increased as compared with the conventional technique using a digital filter.
Of course, the first embodiment described above may be modified so as to switch the threshold value for edge determination, each edge enhancement coefficient, and the reference pattern for attribute determination according to the input image type. Depending on the processing load of each process such as attribute determination, edge detection, and edge enhancement, a modification may be made such that a plurality of processors process in parallel.
In the above-described embodiment, a case has been described in which a program for causing the control unit 210 to execute image processing characteristic of the image processing apparatus according to the present invention is stored in advance in the nonvolatile storage unit 230a. However, for example, the program is recorded and distributed on a computer-readable recording medium such as a CD-ROM (Compact Disk-Read Only Memory) and a DVD (Digital Versatile Disk), and the recording medium is generally used. Of course, the program may be installed in the computer device. In this way, there is an effect that a general computer device can function as the image processing device according to the present invention. Further, in the above-described embodiment, a case has been described in which each characteristic unit of the image processing apparatus according to the present invention is realized by a software module. However, it is needless to say that it may be realized by a hardware module.

(B. Second Embodiment)
Next, an image processing apparatus 500 according to the second embodiment of the present invention will be described.
(B-1: Configuration of the image processing apparatus 500)
FIG. 7 is a functional block diagram illustrating a functional configuration of the image processing apparatus 500.
As illustrated in FIG. 7, the image processing apparatus 500 includes a pre-stage color conversion processing unit 510, a wavelet conversion unit 520, a lightness component processing unit 530, a saturation component processing unit 540, an inverse wavelet conversion unit 550, and a post-stage. A color conversion processing unit 560.

  The pre-stage color conversion processing unit 510 is connected to the image reading apparatus 100 via a communication line such as a USB cable. The pre-stage color conversion processing unit 510 stores a pre-stage color conversion process LUT, and performs the above-described pre-stage color conversion process on the RGB image data received from the image reading apparatus 100 via the communication line. a * b * image data is generated and delivered to the wavelet transform unit 520.

  The wavelet transform unit 520 performs discrete wavelet transform on the L * a * b * image data delivered from the pre-stage color conversion processing unit 510, and the lightness component (L) and the saturation component (a, b) are described above. The wavelet coefficients of the LL component, LH component, HL component, and HH component are generated. Then, the wavelet transform unit 520 delivers each wavelet coefficient generated for the lightness component to the lightness component processing unit 530, and passes each wavelet coefficient generated for the saturation component to the saturation component processing unit 540.

  The lightness component processing unit 530 performs adaptive filter processing, which will be described later, on each wavelet coefficient for the lightness component delivered from the wavelet transform unit 520 and delivers it to the inverse wavelet transform unit 550. The saturation component processing unit 540 performs smoothing processing and saturation conversion processing, which will be described later, on each wavelet coefficient for the saturation component delivered from the wavelet transformation unit 520, and delivers it to the inverse wavelet transformation unit 550. The lightness component processing unit 530 and the saturation component processing unit 540 will be described in detail later.

  The inverse wavelet transform unit 550 receives each wavelet coefficient for the adaptive filter processed lightness component delivered from the lightness component processing unit 530, and the smoothing processing and chroma conversion processing passed from the saturation component processing unit 540. Each wavelet coefficient for the saturation component is subjected to inverse wavelet transform to restore the L * a * b * image data and delivered to the subsequent color conversion processing unit 560.

  The post-stage color conversion processing unit 560 is connected to the image forming apparatus 300 via a communication line such as a USB cable. The post-stage color conversion processing unit 560 stores an LUT for the post-stage color conversion process, and performs the above-described post-stage color conversion process on the L * a * b * image data delivered from the inverse wavelet conversion unit 550. The obtained YMCK image data is delivered to the image forming apparatus 300 via the communication line.

  As described above, the image processing apparatus 500 according to the present embodiment also converts the input RGB image data into L * a * b * image data and performs wavelet conversion, as in the image processing apparatus 200 described above. After that, various processing such as adaptive filter processing, smoothing processing, and saturation conversion processing is performed, and then inverse wavelet transform is performed to restore L * a * b * image data. The image data is further converted into YMCK image data. It is converted and output. Among the components of the image processing apparatus 500, the processes performed by the pre-stage color conversion processing section 510, the wavelet conversion section 520, the inverse wavelet conversion section 550, and the post-stage color conversion processing section 560 are the pre-stage performed by the image processing apparatus 200 described above. Since it is the same as the color conversion process, the wavelet conversion process, the inverse wavelet conversion process, and the subsequent color conversion process, a detailed description thereof will be omitted.

(B-2: Functional configuration of lightness component processing unit 530 and saturation component processing unit 540)
FIG. 8 is a functional block diagram illustrating a functional configuration of the lightness component processing unit 530.
As shown in FIG. 8, the lightness component processing unit 530 includes an edge detection unit 532 and an adaptive filter processing unit 534, and among the wavelet coefficients for the lightness component delivered from the wavelet transform unit 520, the LH component, The HL component and the wavelet coefficients of the HH component are delivered to the edge detection unit 532 and the adaptive filter processing unit 534. Note that the wavelet coefficients of the LL component are delivered to the inverse wavelet transform unit 550 without being subjected to special processing. This is because the lightness component processing unit 530 mainly adjusts the edge, and the edge characteristic does not appear strongly in the LL component.

  The edge detection unit 532 determines whether each pixel constituting the brightness component image is a pixel of interest and determines whether the pixel corresponds to the edge. If the edge detection unit 532 determines that the pixel is an edge, the edge detection unit 532 applies an adaptive filter to a predetermined edge detection signal. In addition to outputting to the processing unit 534 and the saturation component processing unit 540, the edge direction of the edge is detected, and an edge direction signal corresponding to the detection result is output to the saturation component processing unit 540. However, although details will be described later, the edge detection unit 532 determines that the pixel of interest is a halftone dot component even when it is determined that the pixel corresponds to an edge. Edge signals and edge direction signals are not output. Hereinafter, as illustrated in FIG. 9, edge detection processing performed by the edge detection unit 532 with the central pixel (FIG. 9: pixel M) of a pixel block as a target pixel in a 5 × 5 matrix is described with reference to FIG. 10. To do.

FIG. 10 is a flowchart showing the flow of edge detection processing executed by the edge detection unit 532.
As shown in FIG. 10, the edge detection unit 532 first calculates edge continuity for each component of LH (horizontal direction), HL (vertical direction), and HH (slick direction) for the pixel of interest (step SC100). .
More specifically, the edge detection unit 532 sets a target region including the target pixel for each of the LH, HL, and HH components, and calculates an average value of wavelet coefficients in the region. For example, for the LH component, the edge detection unit 532 includes five pixels arranged in the horizontal direction (main scanning direction) around the target pixel in the pixel block shown in FIG. 9 (that is, the pixel K, FIG. 9). L, M, N, and O) are set as target areas. For the HL component, the edge detection unit 532 includes five pixels arranged in the vertical direction (sub-scanning direction) around the target pixel in the pixel block shown in FIG. 9 (that is, the pixel C, FIG. 9). H, M, R, and W) are set as target areas. For the HH component, the edge detection unit 532 includes five pixels arranged in the 45 ° and 135 ° directions with respect to the main scanning direction around the target pixel in the pixel block shown in FIG. For the former, the pixels E, I, M, Q, and U of FIG. 9 are set as the target region, and for the latter, the pixels A, G, M, S, and Y) are set as the target regions. Next, the edge detection unit 532 calculates an absolute value of a difference between the pixel value and the average value for each pixel in the target region, and a value obtained by multiplying the absolute value by a predetermined gain coefficient F. The number of pixels that exceeds a predetermined threshold is calculated as the edge continuity.

  The gain coefficient F is a value having the correspondence relationship shown in FIG. 10 with the background pixel density value bmin. As shown in FIG. 10, the gain coefficient F is “0” until the background pixel density value bmin reaches Dth1, and increases linearly when the background pixel density value bmin exceeds Dth1, and the background pixel density After the value bmin reaches Dth2, it has a characteristic that it becomes a constant value (1.0). This gain factor F makes it easy to detect an edge in an area where the background pixel density of the LL component is low, such as a character area, when detecting an edge in the HL, LH and HH components for the brightness component. This is to make it difficult to detect the gain of an area having a high background pixel density such as an area. The background pixel density value bmin is a value obtained as follows. That is, in the pixel block shown in FIG. 9, an average value of pixel values of LL components for five pixels (for example, pixel A to pixel E) arranged in the main scanning direction (row direction) is obtained for each row. The average value of the pixel values of the LL component is obtained for each column for five pixels (for example, pixel A to pixel U) arranged in the sub-scanning direction (column direction), and the smallest value among these ten average values Is the background pixel density value bmin. In this embodiment, the gain coefficient calculation LUT representing the correspondence shown in FIG. 10 is stored in the edge detection unit 532 in advance, and the gain coefficient F corresponding to the background pixel density value bmin calculated as described above is obtained. By reading from the gain coefficient calculation LUT, the gain coefficient F corresponding to the target pixel is acquired.

  In step SC110 executed subsequent to step SC100, the edge detection unit 532 determines the edge continuity calculated in step SC100 for each of the four directions of the horizontal direction, the vertical direction, the 45 ° direction, and the 135 ° direction. To determine whether or not at least one of the four edge continuity values exceeds the threshold value. If at least one of the four edge continuity values exceeds the threshold value (step SC110: Yes), the edge detection unit 532 determines whether the pixel of interest is a halftone dot component. (Step SC120). Conversely, if none of the four edge continuity values exceeds the threshold value, the edge detection unit 532 determines that the pixel of interest is not an edge, and ends the edge detection process.

  In step SC120, the edge detection unit 532 uses the ten variables Ea1, Ea2, Ea3, Ea4, Eb1, Eb2, Eb3, Eb4, aa, and bb to configure the pixel of interest as a halftone dot. Determine if it is an element. Specifically, in the edge detection unit 532, first, “0” or “1” is set in each of Ea1 to Ea4 and Eb1 to Eb4 as described below.

(1) Ea1 to Ea4 are set with reference to wavelet coefficients (pixel values) of HH components.
Specifically, the edge detection unit 532
When the absolute value of the difference between the pixel value of the pixel A and the pixel value of the target pixel (that is, the pixel M) exceeds a predetermined threshold, “1” is set to Ea1, and in the opposite case, Ea1 Set “0” to
When the absolute value of the difference between the pixel value of the pixel E and the pixel value of the target pixel exceeds a predetermined threshold, “1” is set to Ea2, and in the opposite case, “0” is set to Ea2. And
When the absolute value of the difference between the pixel value of the pixel U and the pixel value of the target pixel exceeds a predetermined threshold, “1” is set to Ea3, and in the opposite case, “0” is set to Ea3. And
When the absolute value of the difference between the pixel value of the pixel Y and the pixel value of the target pixel exceeds a predetermined threshold, “1” is set to Ea4, and in the opposite case, “0” is set to Ea4. To do.

(2) Eb1 and Eb2 are set with reference to the wavelet coefficient of the HL component.
Specifically, the edge detection unit 532
When the absolute value of the difference between the pixel value of the pixel C and the pixel value of the target pixel exceeds a predetermined threshold value, “1” is set to Eb1, and in the opposite case, “0” is set to Eb1. And
When the absolute value of the difference between the pixel value of the pixel W and the pixel value of the target pixel exceeds a predetermined threshold, “1” is set to Eb2, and in the opposite case, “0” is set to Eb2. To do.

(3) Eb3 and Eb4 are set with reference to the wavelet coefficients of the LH component.
Specifically, the edge detection unit 532
When the absolute value of the difference between the pixel value of the pixel K and the pixel value of the target pixel exceeds a predetermined threshold, Eb3 is set to “1”, and in the opposite case, Eb3 is set to “0”. And
When the absolute value of the difference between the pixel value of the pixel O and the pixel value of the pixel of interest exceeds a predetermined threshold, “1” is set to Eb4, and in the opposite case, “0” is set to Eb4. To do.

Then, the edge detection unit 532 sets the sum of Ea1 to Ea4 to aa and sets the sum of Eb1 to Eb4 to bb.
When the value of aa is “4” and the value of bb is “0”,
Or, when the value of aa is “0” and the value of bb is “4”,
Alternatively, if both aa and bb are both “4”, it is determined that the target pixel is a halftone dot component.
The above is the halftone dot determination process executed by the edge detection unit 532 in step SC120 shown in FIG. In the present embodiment, the case where it is determined whether the pixel of interest is a halftone dot component using the above ten variables has been described. However, the determination may be performed by another method. Of course.

  As shown in FIG. 10, when the determination result in step SC120 is “Yes” (that is, when it is determined that the pixel of interest is a component of halftone dots), the edge detection unit 532 performs the processing as described above. The edge detection processing is terminated without outputting the edge detection signal. Conversely, when the determination result in step SC120 is “No”, the edge detection signal is output (step SC130).

  In step SC140 executed subsequent to step SC130 in FIG. 10, the edge detection unit 532 detects the edge direction of the target pixel determined to be an edge and outputs an edge direction detection signal representing the detection result. To do. More specifically, the edge detection unit 532 detects the edge direction using six variables Ex1, Ex2, Ey1, Ey2, Lx, and Ly.

Specifically, the edge detection unit 532 first sets values for each of Ex1, Ex2, Ey1, and Ey2 as described below.
For Ex1, the sum of the pixel values of the pixel G, pixel H and pixel I (see FIG. 9) of the LH component is set.
For Ex2, the sum of the pixel values of the LH component pixel Q, pixel R and pixel S (see FIG. 9) is set.
For Ey1, the sum of the pixel values of the HL component pixel G, pixel L, and pixel Q (see FIG. 9) is set.
For Ey2, the sum of pixel values of the HL component pixel I, pixel N, and pixel S (see FIG. 9) is set.

Then, the edge detection unit 532 sets the absolute value of the difference between Ex1 and Ex2 calculated as described above to Lx, and sets the absolute value of the difference between Ey1 and Ey2 to Ly.
Next, the edge detection unit 532 compares Lx and Ly with a predetermined threshold value, and if both are smaller than the threshold value, an edge indicating that there is no edge direction (that is, no edge). When a direction signal is output and at least one of Lx and Ly exceeds the threshold value, it is determined which of Lx and Ly is greater.

  When Lx is larger than Ly, the edge detection unit 532 further determines which of the above-described Ex1 and Ex2 is larger, and when Ex1 is large, the pixel region used when calculating Ex1 (that is, Ex1) , The direction along the pixels G, H, and I (hereinafter referred to as direction 1) is determined to be the edge direction, and an edge direction signal representing the direction is output. Conversely, when Ex2 is large, It is determined that the direction (hereinafter referred to as direction 2) along the pixel region (that is, pixels Q, R, and S) used when calculating Ex2 is the edge direction, and an edge direction signal indicating the direction is output. To do.

Conversely, when Ly is greater than or equal to Lx, the edge detection unit 532 further determines which of the above-described Ey1 and Ey2 is greater, and when Ey1 is greater, the pixel used when calculating Ey1 When the direction along the region (that is, the pixels G, L, and Q) (hereinafter referred to as direction 3) is determined to be the edge direction, an edge direction signal indicating the direction is output, and conversely, Ey2 is large Determines that the direction (hereinafter referred to as direction 4) along the pixel region (that is, pixels I, N, and S) used in calculating Ey2 is the edge direction, and represents the edge direction. Output a signal.
The above is the edge direction detection processing executed by the edge detection unit 532 in step SC140 shown in FIG. Although details will be described later, the edge direction signal output from the edge detection unit 532 in this way is a ghost (for example, an edge of another color such as red appears in the vicinity of the black edge, and the color blurs at the edge part). This is used when the saturation component processing unit 540 executes a smoothing process for correcting the phenomenon that occurs. In the present embodiment, the case where the edge direction is detected using the above six variables has been described, but it is needless to say that the edge direction may be detected by other methods.

The adaptive filter processing unit 534 multiplies the wavelet coefficients of the LH component, the HL component, and the HH component corresponding to the pixel of interest by any one of the above-described character edge enhancement coefficient and pattern edge enhancement coefficient, and outputs an inverse wavelet. The data is output to the conversion unit 550. More specifically, when an edge detection signal is output from the edge detection unit 532 for the pixel of interest, the wavelet coefficients are multiplied by a character edge enhancement coefficient and output. When the detection signal is supplied, it is multiplied by the pattern edge enhancement coefficient and output. Thereby, filter processing in real space is realized.
The above is the configuration and function of the lightness component processing unit 530.

(B-3: Functional configuration of the saturation component processing unit 540)
FIG. 12 is a functional block diagram illustrating a functional configuration of the saturation component processing unit 540.
As illustrated in FIG. 12, the saturation component processing unit 540 includes a smoothing processing unit 542, a saturation calculation unit 544, a black edge strength calculation unit 546, and a saturation conversion unit 548. Note that the smoothing processing unit 542 and the saturation conversion unit 548 are actually provided for each saturation component (that is, for each a * component and b * component), but in order to simplify the description. In FIG. 12, one by one is illustrated.

  Of the wavelet coefficients for the saturation component delivered from the wavelet transform unit 520, the LL component is delivered to the saturation calculation unit 544 and the saturation transformation unit 548 corresponding to the saturation component for each saturation component, and the LH The component, the HL component, and the HH component are delivered to the smoothing processing unit 542 corresponding to the saturation component for each saturation component.

  The smoothing processing unit 542 sets the wavelet coefficients of the LH component, the HL component, and the HH component of the saturation components output from the wavelet transform unit 520 for each saturation component (ie, a * component, b *). For each component) and the aforementioned edge direction signal is delivered. The smoothing processing unit 542 rewrites each wavelet coefficient in a region corresponding to the signal value of the edge direction signal (hereinafter referred to as a smoothing region) to a predetermined value (in this embodiment, “0”) and smoothes the smoothed signal. The saturation of the region is smoothed (that is, black is changed to one color) and output to the inverse wavelet transform unit 550.

  As described above, the edge direction detection signal output from the edge detection unit 532 can take five types of signal values: a signal value indicating no edge direction and a signal value indicating any one of the directions 1 to 4. . In the present embodiment, the smoothing processing unit 542 sets a smoothing region as follows for the above five types of signal values of the edge direction detection signal. That is, when the signal value indicates that there is no edge, a 3 × 3 matrix area centered on the target pixel (FIG. 9: pixel M) is selected as the smoothing area. In addition, when the signal value indicates the direction 1, the smoothing processing unit 542 selects the region including three pixels corresponding to the direction 2 described above as the smoothing region, and the signal value indicates the direction 2. In this case, a region composed of three pixels corresponding to the above-described direction 1 is selected as the smoothing region. Then, when the signal value indicates the direction 3, the smoothing processing unit 542 selects the region including the three pixels corresponding to the direction 4 described above as the smoothing region, and the signal value indicates the direction 4. In this case, a region composed of three pixels corresponding to the above-described direction 3 is selected as the smoothing region. Since the pixel values (wavelet coefficients) of the LH, HL, and HH components corresponding to the smoothing area selected in this way are all rewritten to “0”, the smoothing area is converted to a black one-color area. Generation of ghosts is suppressed.

The saturation calculation unit 544 receives the wavelet transform coefficient (only the LL component) of each saturation component, calculates the saturation C * according to the wavelet coefficient, and delivers it to the black edge intensity calculation unit 546. More specifically, the saturation calculation unit 544 performs 4-bit quantization on the wavelet coefficient for the a * component and the wavelet coefficient for the b * component, respectively, and then squares each value. Addition is performed, and the square root of the addition value is output as C *.
In this embodiment, the case where the saturation C * is calculated by executing the above calculation has been described. However, the value that can be taken by the wavelet coefficient for the a * component and the value that can be taken by the wavelet coefficient for the b * component. In addition, the saturation calculation unit 544 stores the LUT in which the value is used in association with the value of the saturation C * calculated by the above calculation, and the saturation C * corresponding to the input wavelet coefficient is stored. Of course, it may be obtained by referring to the LUT. In this way, it is not necessary to perform saturation calculation in the image processing execution process, and the processing speed can be increased.

The black edge strength calculation unit 546 calculates the saturation conversion coefficient fc based on the saturation C * delivered from the saturation calculation unit 544 and also uses the edge detection signal delivered from the edge detection unit 532 to determine the edge strength. e is calculated, and an edge strength coefficient fe is calculated based on the edge strength e. Here, the edge strength e is the absolute value of the wavelet coefficient having the maximum value among the wavelet coefficients of the LH, HL, and HH components corresponding to the target pixel. Then, the black edge strength calculation unit 546 calculates 1.0 by adding 1.0 to the product of the saturation conversion coefficient fc and the edge strength coefficient fe, and calculates the black edge strength fb as the saturation conversion unit 548. Output to.
More specifically, the black edge strength calculation unit 546 includes an edge strength coefficient LUT representing the correspondence relationship shown in FIG. 13A for the edge strength e and the edge strength coefficient fe, and the saturation c * and the saturation. A saturation conversion coefficient LUT representing the correspondence shown in FIG. 13B is stored for the conversion coefficient fc. The black edge strength calculation unit 546 obtains the saturation conversion coefficient fc and the edge strength coefficient fe with reference to these LUTs, performs the above calculation, and calculates the black edge strength fb. The LUT in which the saturation conversion coefficient fc and the edge intensity coefficient fe are associated with the value of the black edge intensity fb calculated by the combination thereof is stored in the black edge intensity calculation unit 546, and the LUT is referred to. Of course, the black edge strength fb may be acquired.

The saturation conversion unit 548 multiplies the wavelet coefficients a * and b * by the black edge strength fb delivered from the black edge strength calculation unit 546, and delivers the calculation result to the inverse wavelet transformation unit 550. Here, the black edge strength fb is a combination of whether the pixel of interest belongs to an edge portion or a non-edge portion, and whether the pixel of interest belongs to a low saturation (ie, black) region or a high saturation region. Accordingly, since the values shown in FIG. 13C are obtained, the wavelet coefficients a * ′ and b * ′ after the conversion by the saturation conversion unit 548 become the values shown in FIG. The point to be noted here is that when the target pixel belongs to the low saturation region and the edge region, the values of a * ′ and b * ′ are both 0 (that is, pure black). It is. This avoids the occurrence of a ghost near the black edge.
The above is the configuration and function of the saturation component processing unit 540.

(B-4: Effects of the second embodiment)
As described above, since the image processing apparatus 500 according to the second embodiment is configured as described above, the number of processing dimensions is smaller than when image processing such as edge detection is performed without performing wavelet transform. Can be greatly reduced, and the processing speed is improved.
In addition, according to the present embodiment, since filter processing such as edge enhancement is performed by coefficient multiplication on wavelet coefficients, it is compared with the conventional technique that performs filter processing by product-sum operation of pixel values using a digital filter. Thus, it is possible to reduce the amount of calculation required for the filter processing.
In the second embodiment described above, the image processing apparatus according to the present invention includes a pre-stage color conversion processing unit 510, a wavelet conversion unit 520, a lightness component processing unit 530, a saturation component processing unit 540, an inverse wavelet conversion unit 550, and The case where the subsequent color conversion processing unit 560 is configured in combination has been described. For example, a program that realizes the functions of the above units in the computer device having the hardware configuration illustrated in FIG. 2 is stored in the nonvolatile storage unit. You may make it install and operate a control part according to the program.

(C: Modification)
Although the first and second embodiments of the present invention have been described above, it goes without saying that the following modifications may be added to such embodiments.
(C-1: Modification 1)
In the first and second embodiments described above, the case where the RGB image data delivered from the image reading apparatus 100 is converted into YMCK image data and delivered to the image forming apparatus 300 has been described. However, the output destination of the processing result is an image. The present invention is not limited to the forming apparatus, and the processing result may be delivered to a computer device having a display device such as a liquid crystal display, and an image corresponding to the processing result may be displayed on the display device. Alternatively, the processing result may be delivered to a computer apparatus provided with a database and registered in the database.

(C-2: Modification 2)
In the first and second embodiments described above, the case where the RGB image data delivered from the image reading apparatus 100 is converted into L * a * b * image data by the pre-stage color conversion processing and then wavelet conversion is performed has been described. Of course, wavelet transform may be applied after conversion to YCbCr image data. In short, any color system may be used as long as wavelet transform is performed after the input image data is converted into image data representing an image with a brightness component and a saturation component. In the first and second embodiments described above, the case where the input image data is RGB image data and the output image data is YMCK image data has been described. However, both of the input image data and the output image data are different. Of course, it may be image data of the color system. When the input image data is image data representing an image with lightness components and saturation components (for example, L * a * b * image data or YCbCr image data), it is necessary to perform pre-stage color conversion processing. In addition, when the output image data is image data representing an image with a lightness component and a saturation component, it is not necessary to perform post-stage color conversion processing.

(C-3: Modification 3)
In the first and second embodiments described above, the case where the wavelet transform and the inverse wavelet transform are performed using the Haar filter has been described. However, the filter used for the wavelet transform is switched according to the property of the image to be processed. Also good.

(C-4: Modification 4)
In the first embodiment described above, the case has been described in which either the character edge enhancement coefficient or the design edge enhancement coefficient is multiplied by the wavelet coefficient corresponding to the image block in accordance with the attribute of the image block. However, if the attribute of an image block is determined to be, for example, n% as a character probability and m% as a pattern according to the degree of deviation from each reference pattern, it corresponds to the image block. An edge enhancement coefficient to be multiplied by the wavelet coefficient is generated by adding a value obtained by multiplying the character edge enhancement coefficient by n / 100 and a value obtained by multiplying the pattern edge enhancement coefficient by m / 100. Also good. Similarly, for the edge enhancement coefficient used in the adaptive filter processing unit 534 of the image processing apparatus 500 according to the second embodiment, the character edge enhancement coefficient and the pattern edge enhancement coefficient are added at a ratio corresponding to the edge strength. You may make it produce | generate.

(C-5: Modification 5)
In the first embodiment described above, the case where the attribute unknown block is used by switching between the character edge enhancement coefficient and the pattern edge enhancement coefficient according to the edge strength of the edge portion has been described. You may make it perform the edge emphasis process which the image processing apparatus 500 which concerns on 2nd Embodiment performs.

It is a figure which shows the structural example of the image processing system containing the image processing apparatus 200 which concerns on one Embodiment of this invention. 2 is a diagram illustrating a configuration example of the image processing apparatus 200. FIG. It is a figure which shows an example of the reference | standard pattern stored in the attribute discrimination | determination table. It is a flowchart which shows the flow of the process of the edge emphasis process which the control part 210 performs. It is a flowchart which shows the flow of the edge process which the control part 210 performs to a character block or a picture block. It is a flowchart which shows the flow of the edge process which the control part 210 performs to an unknown block. It is a functional block diagram which shows the function structure of the image processing apparatus 500 which concerns on 2nd Embodiment of this invention. 5 is a functional block diagram showing a functional configuration of a lightness component processing unit 530. FIG. FIG. 6 is a diagram illustrating an example of a pixel of interest and its peripheral pixels for edge detection processing executed by an edge detection unit 532. 5 is a flowchart showing a flow of edge detection processing executed by the edge detection unit 532; It is a figure which shows an example of the gain function utilized in the edge detection process. 3 is a functional block diagram illustrating a functional configuration of a saturation component processing unit 540. FIG. It is a figure for demonstrating the relationship between edge intensity | strength e, edge conversion coefficient fe, saturation C *, saturation conversion coefficient fc, and black edge intensity | strength fb. It is a figure for demonstrating saturation a * 'and b *' after a saturation conversion process. It is a figure for demonstrating wavelet transformation and its reverse transformation.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 100 ... Image reading apparatus, 200,500 ... Image processing apparatus, 300 ... Image forming apparatus, 210 ... Control part, 220 ... IF part, 230 ... Storage part, 230a ... Nonvolatile storage part, 230b ... Volatile storage part, 240 ... Bus 510 ... Previous color conversion processing unit 520 ... Wavelet conversion unit 530 ... Brightness component processing unit 532 ... Edge detection unit 534 ... Adaptive filter processing unit 540 ... Saturation component processing unit 542 ... Smoothing processing , 544... Saturation calculation unit, 546... Black edge strength calculation unit, 548. Saturation conversion unit, 550. Inverse wavelet conversion unit, 560.

Claims (5)

Input means for inputting image data representing an image with brightness and saturation of each pixel;
Wavelet transform means for performing wavelet transform on the image data, and generating wavelet coefficients of LL component, LH component, HL component and HH component for each of the lightness component and saturation component of the image represented by the image data;
Discrimination that divides an image represented by the wavelet coefficient of the LL component for the lightness component into a plurality of blocks, and determines whether or not the attribute is a background with reference to the wavelet coefficient of the corresponding LL component for each of the blocks Means,
For the block determined by the determining means as a background, the wavelet coefficients of the HL component, the LH component, and the HH component corresponding to the block are rewritten to predetermined values, while the block is determined as not being a background. An edge processing means for detecting the edge portion with reference to the wavelet coefficients of the HL component, the LH component, and the HH component corresponding to the block, and multiplying the block by a predetermined edge enhancement coefficient;
Image data is restored by performing inverse wavelet transform on the wavelet coefficients of the LL component and the wavelet coefficients of the HL component, the LH component, and the HH component that have been processed according to the attribute by the edge processing means, An output means for outputting data;
An image processing apparatus comprising: The discrimination means includes
For a block that is determined not to be a groundwork, determine whether the attribute is text, design, or unknown,
The edge processing means includes
For a block whose attribute is determined to be a character by the determination means, the wavelet coefficient of the HL component, LH component and HH component corresponding to the edge portion is multiplied by a first edge enhancement coefficient,
For the block whose attribute is determined to be a character by the determination means, the second wavelet coefficient corresponding to the edge portion is smaller than the first edge enhancement coefficient in the second wavelet coefficient of the HL component, LH component, and HH component. Multiply the edge enhancement factor of
For the block whose attribute is determined to be unknown, each of the wavelet coefficients of the HL component, the LH component, and the HH component corresponding to the edge portion is compared with a predetermined threshold value, and is larger than the threshold value. 2. The image processing apparatus according to claim 1, wherein the first edge enhancement coefficient is multiplied in the case, and the second enhancement coefficient is multiplied in the case of being smaller than the threshold value. Input means for inputting image data representing an image with brightness and saturation of each pixel;
Wavelet transform that performs wavelet transform on the image data input to the input means and generates wavelet coefficients of LL component, LH component, HL component, and HH component for each of the brightness component and saturation component of the image represented by the image data Means,
A lightness component that detects the strength and direction of an edge with reference to the wavelet coefficient of the lightness component, and multiplies the wavelet coefficients of the LH component, the HL component, and the HH component corresponding to the edge by an edge enhancement coefficient corresponding to the strength. Processing means;
Among the wavelet coefficients of the saturation component, the LL component has a predetermined wavelet coefficient corresponding to the edge according to the saturation calculated from the LL component and the intensity of the edge detected by the lightness component processing means. While performing the saturation conversion by multiplying the values, for the LH component, the HL component, and the HH component, the values of the wavelet coefficients corresponding to the region determined according to the edge direction detected by the lightness component processing means are set to predetermined values. Saturation component processing means for rewriting to
An output unit for performing inverse wavelet transform on the processing result by the lightness component processing unit and the processing result by the saturation component processing unit to restore image data, and outputting the image data;
An image processing apparatus comprising: Computer equipment,
Image data representing an image with lightness and saturation of each pixel is received, wavelet transform is performed on the image data, and LL component, LH component, HL for each of the lightness component and saturation component of the image represented by the image data A first step of generating component and HH component wavelet coefficients;
The image represented by the wavelet coefficient of the LL component for the lightness component is divided into a plurality of blocks, and for each of them, it is determined whether or not the attribute is a background by referring to the wavelet coefficient of the corresponding LL component. Two steps,
For the block determined as the background in the second step, the wavelet coefficients of the HL component, the LH component, and the HH component corresponding to the block are rewritten to predetermined values. For the block determined in step 2, a third edge is detected by referring to the wavelet coefficients of the HL component, LH component, and HH component corresponding to the block, and multiplying by a predetermined edge enhancement coefficient. Steps,
Image data is restored by performing inverse wavelet transform on the wavelet coefficients of the LL component and the wavelet coefficients of the HL component, the LH component, and the HH component that have been processed according to the attribute in the third step, A fourth step of outputting the image data;
A program characterized by having executed. Computer equipment,
Image data representing an image with lightness and saturation of each pixel is received, wavelet transform is performed on the image data, and LL component, LH component, HL for each of the lightness component and saturation component of the image represented by the image data A first step of generating component and HH component wavelet coefficients;
The edge strength and direction are detected with reference to the wavelet coefficient of the lightness component, and the LH component, the HL component, and the HH component wavelet coefficients corresponding to the edge are multiplied by an edge enhancement coefficient corresponding to the strength. And the steps
Among the wavelet coefficients of the saturation component, for the LL component, the wavelet coefficient corresponding to the edge is predetermined according to the saturation calculated from the LL component and the intensity of the edge detected in the second step. In the meantime, for the LH component, the HL component, and the HH component, wavelet coefficient values corresponding to regions determined in accordance with the edge direction detected in the second step are predetermined. A third step of rewriting to the value of
A fourth step of performing inverse wavelet transform on the processing result of the second step and the processing result of the third step to restore the image data, and outputting the image data;
A program characterized by having executed.

Images