JP5017198B2 - Image processing apparatus and image processing program - Google Patents

Description

  The present invention relates to an image processing apparatus and an image processing program for performing encoding after wavelet transform of an input image.

  In recent years, with the development of digital devices, various compression methods for image data and the like have been proposed, and JPEG (Joint Photographic Experts Group) 2000 is known as one of the compression methods. JPEG2000 defines an image compression / decompression method, and is an extension of JPEG. JPEG2000 divides image data into a plurality of tiles, and generates encoded data by performing processing such as wavelet transform, quantization, coefficient bit modeling, and arithmetic coding for each tile.

However, when JPEG2000 is used, there is a problem that the compression efficiency of an image composed of a periodic pattern (halftone dot) such as a halftone image is poor. Therefore, Patent Document 1 discloses a method of comparing an input image with a pattern image that has already been stored and compressing the pattern image if they match to improve the compression rate. Furthermore, Patent Document 2 describes a method for improving the compression ratio by encoding and compressing the difference between a run length for counting the number of consecutive identical data and adjacent pixels.
Japanese Patent No. 3862637 JP-A-11-177828

  However, in the case of the method described in Patent Document 1, it is necessary for the input image and the pattern image to completely coincide with each other, and it is difficult to adapt to a pattern such as a halftone image. Further, in the case of the method described in Patent Document 2, the compression efficiency is not improved because the same data is rarely continuous for halftone images.

  The present invention has been made to solve the above problems, and an object thereof is to provide an image processing apparatus and an image processing program for improving the compression ratio for an image having a periodic pattern such as a halftone image. To do.

An image processing apparatus according to the present invention includes a difference image acquisition unit that acquires a difference image between an input image and an approximate pattern image of the input image, and an approximation that stores the approximate pattern image and image information of the difference image in association with each other. Pattern image storage means; wavelet transform means for performing wavelet transform on the difference image to output coefficient data; quantization means for quantizing the coefficient data to output quantized data; and encoding the quantized data Encoding means for outputting encoded data, frequency component detection means for detecting a frequency component of the input image, and pattern image generation means for generating the approximate pattern image based on the detected frequency component, The difference image acquisition means acquires a difference image between the input image and the generated approximate pattern image.

The image processing program according to the present invention includes a difference image acquisition unit that acquires a difference image between an input image and an approximate pattern image of the input image, a wavelet conversion unit that performs wavelet conversion on the difference image and outputs coefficient data, Quantizing means for quantizing the coefficient data and outputting quantized data, encoding means for encoding the quantized data and outputting encoded data, frequency component detecting means for detecting a frequency component of the input image, It functions as a pattern image generation unit that generates the approximate pattern image based on the detected frequency component, and the difference image acquisition unit acquires a difference image between the input image and the generated approximate pattern image.

  According to these configurations, it is possible to easily generate an approximate pattern image of the input image by measuring the frequency characteristics of the input image, and the difference image between the input image and the approximate pattern image is higher frequency component and lower frequency component than the input image. Therefore, the compression rate can be improved.

The image processing apparatus according to the present invention stores difference image acquisition means for acquiring a difference image between an input image and an approximate pattern image of the input image, and stores the approximate pattern image and image information of the difference image in association with each other. An approximate pattern image storage means, a wavelet transformation means for performing wavelet transformation on the difference image and outputting coefficient data, a quantization means for quantizing the coefficient data and outputting quantized data, and the quantized data Encoding means for encoding and outputting encoded data, pattern image storage means for storing a plurality of pattern images showing different frequency components, and difference for generating a difference image between each of the stored pattern images and the input image An image generation means, wherein the difference image acquisition means obtains a difference image with the smallest data amount among the generated difference images. -Option to acquire.

In the image processing program according to the present invention, the computer associates the difference image acquisition means for acquiring a difference image between the input image and the approximate pattern image of the input image, and associates the approximate pattern image with the image information of the difference image. Approximate pattern image storage means for storing, wavelet transform means for performing wavelet transform on the difference image and outputting coefficient data, quantizing means for quantizing the coefficient data and outputting quantized data, encoding the quantized data Encoding means for generating encoded data and outputting encoded data, pattern image storage means for storing a plurality of pattern images showing different frequency components, difference image generating means for generating a difference image between each of the stored pattern images and an input image And the difference image acquisition means has the largest amount of data among the generated difference images. To get to select the less the difference image.

  According to these configurations, an approximate pattern image of an input image can be easily obtained by storing in advance a plurality of pattern images having different frequencies in the pattern image storage means. Further, since the difference image between the input image and the approximate pattern image has less data of high frequency components and low frequency components than the input image, the compression rate can be improved.

  Here, it is desirable to investigate in advance the frequency characteristics of a pattern that causes a particularly poor compression rate, and to store many pattern images that can be applied to an input image having the frequency characteristics. The pattern image storage means stores a plurality of pattern images having different frequency characteristics as table data, and the difference image acquisition means reads the pattern image from the table as appropriate to generate a difference image, and generates the largest amount of data. The file name, pattern number, and the like of the pattern image that generated a small number of difference images may be added to the compressed difference image and stored. In addition to storing a plurality of pattern images as table data, the pattern image storage means may have a basic expression that approximates a frequency component, and may store the coefficients thereof.

Further, in the above-described configuration, an inverse encoding means for inversely encoding the encoded data and outputting quantized data, an inverse quantization means for inversely quantizing the quantized data and outputting coefficient data, and the coefficient data An inverse wavelet transform unit that performs inverse wavelet transform to output a differential image, and reads out an approximate pattern image in which image information of the differential image is stored in association with the approximate pattern image storage unit, and stores the differential image and restoring means for restoring said input image, further comprising preferably Rukoto a.

  According to this configuration, the decoded difference image can be easily restored to the input image by reading out the approximate pattern image stored in the approximate pattern image storage unit.

  According to the present invention, the approximate pattern image of the input image can be easily generated by measuring the frequency characteristics of the input image, and the difference image between the input image and the approximate pattern image has a higher frequency component and a lower frequency component than the input image. Since the data is reduced, the compression rate can be improved.

  Further, by storing a plurality of pattern images having different frequencies in advance, an approximate pattern of the input image can be easily obtained. Further, since the difference image between the input image and the approximate pattern image has less data of high frequency components and low frequency components than the input image, the compression rate can be improved.

<First Embodiment>
Embodiments of an image processing apparatus and an image processing program according to the present invention will be described with reference to the drawings. FIG. 1 is a block diagram showing an electrical configuration of an image processing apparatus 1 in the present embodiment. The image processing apparatus 1 is realized by an information processing apparatus such as a personal computer (personal computer), a workstation, or a portable information terminal, and the block diagram of FIG. 1 shows a case where the image processing apparatus 1 is realized by these information processing apparatuses. Is shown as an example. In addition, it may be used by being incorporated in an image reading apparatus such as a copying machine or a scanner, or an imaging apparatus such as a digital camera, a digital video camera, or a mobile phone equipped with a digital camera.

  As shown in FIG. 1, the image processing apparatus 1 includes a control unit 11, a storage unit 12, an input operation unit 13, a display unit 14, an I / F unit 15, a network I / F unit 16, an image processing unit 17, and the like. Composed. The control unit 11 is configured by a CPU (Central Processing Unit) or the like, reads a program stored in the storage unit 12 in accordance with an input instruction signal or the like, executes a process, and sends the processing to each functional unit. The image processing apparatus 1 is comprehensively controlled by outputting an instruction signal, transferring data, and the like.

  The storage unit 12 stores programs, data, and the like for realizing various functions provided in the image processing apparatus 1. The input operation unit 13 includes various operation buttons and a pointing device such as a mouse, and outputs an operation signal to the control unit 11 when operated by a user. The user also sets the compression rate via the input operation unit 13. The display unit 14 is a display screen such as a liquid crystal display, and displays according to the content input from the input operation unit 13 or displays the processing content and processing result by the control unit 11.

  The I / F unit 15 is an interface such as IEEE1394 or USB, and can directly transmit / receive data to / from an external device. The network I / F unit 16 includes a communication module such as a LAN board, and transmits / receives various data to / from an external device via a network (not shown) connected to the network I / F unit 16.

  The image processing unit 17 is an image processing IC for performing compression / decoding processing of an input image, and is controlled by the control unit 11. The image processing unit 17 includes a storage unit 21 and a RAM (Random Access Memory) 22. The storage unit 21 stores a compression program 31 and a decoding program 32. The compression program 31 is a program executed by the image processing unit 17 when an input image input via the I / F unit 15 or the network I / F unit 16 is compressed by the JPEG2000 system. The decoding program 32 is a program executed by the image processing unit 17 when decoding the compressed image subjected to the compression processing by the image processing unit 17. The RAM 22 is a working memory of the image processing unit 17 and temporarily stores an approximate pattern image generated by executing the compression program 31.

  FIG. 2 is a flowchart showing the flow of the image compression process performed when the image processing unit 17 reads and executes the compression program 31. 3 and 5 are diagrams for explaining the processing of JPEG2000, and FIG. 4 is a diagram for explaining the generation of a difference image. The flow of compression processing in the present embodiment will be described with reference to FIGS.

  First, the image processing unit 17 performs component conversion of the color system of the input image data, for example, from the RGB system to the YCbCr system (step S11), and a plurality of image data converted from the color system are displayed as shown in FIG. Are divided into tiles (step S12).

  After the tile division, the image processing unit 17 detects the frequency characteristics of the tile image (step S13), and generates an approximate pattern image based on the frequency characteristics (step S14). In FIG. 4, 51 is a tile image, and 52 is an approximate pattern image of a tile image generated based on the frequency characteristics of the tile image. The approximate pattern image 52 is a pattern image having a periodic pattern. Here, when the image processing apparatus 1 can execute JPEG compression in addition to JPEG 2000, the frequency characteristics of the tile image 52 may be measured using discrete cosine transform performed at the time of JPEG compression.

  Then, the image processing unit 17 stores the generated approximate pattern image 52 and the image information of the difference image 55 in association with each other in the RAM 22 (step S15). The approximate pattern image 22 stored in the RAM 22 is used when the compressed image is decoded. The image information of the difference image 55 refers to unique information given to each difference image, and is, for example, position information of the tile image 51 from which the difference image 55 is generated.

  Next, the image processing unit 17 generates a difference image obtained by taking the difference between the tile image 51 and the approximate pattern image 52 (step S16). In FIG. 4, 53 is a difference image. The difference image 53 is an image in which, when the approximate pattern image 52 is superimposed on the tile image 51, a portion where the two images overlap is removed, and a portion where only the components of the tile image 51 remain is not overlapped.

  Then, the image processing unit 17 performs discrete wavelet transform on the difference image 53 and outputs coefficient data (step S17). FIG. 5 is a schematic diagram showing the discrete wavelet transform. A discrete wavelet transform is applied to the tile image (0LL) in FIG. 5A, and the tile image is decomposed into subbands 1LL, 1HL, 1LH, and 1HH as shown in FIG. 5B. Subsequently, a discrete wavelet transform is performed on the low frequency component 1LL, and it is decomposed into subbands 2LL, 2HL, 2LH and 2HH as shown in FIG. Note that the image 54 in FIG. 4 illustrates 1-level conversion and FIG. 5 illustrates 2-level conversion, but the number of discrete wavelet transforms is not particularly limited.

  Subsequently, the image processing unit 17 performs linear quantization on the coefficient data subjected to the discrete wavelet transform (Step S18), and performs entropy coding on the quantized quantized data (Step S19). Since quantization and entropy coding are well known in each document, description thereof is omitted.

  Next, the flow of decoding processing will be described. FIG. 6 is a flowchart showing the flow of the decoding process performed when the image processing unit 17 reads and executes the decoding program 32. First, the image processing unit 17 performs inverse entropy coding on the compressed image (step S21), and acquires the quantized data generated in step S18 of FIG. Subsequently, the image processing unit 17 performs inverse quantization on the quantized data, and obtains coefficient data subjected to the discrete wavelet transform generated in step S17 of FIG. 2 (step S22).

  Then, the image processing unit 17 performs inverse discrete wavelet transform on the coefficient data, and acquires the difference image 53 generated in step S16 of FIG. 2 (step S23). The image processing unit 17 reads the approximate pattern image 51 stored in association with the image information of the difference image 53 stored in the RAM 22, and restores the difference image 53 to the tile image 51 (step S24). Then, the image processing unit 17 performs colorimetric component conversion of the image obtained by combining the tile images, and ends the decoding of the compressed image (step S25).

  When a conventional JPEG2000 compression technique is used to compress an image composed of a periodic pattern (halftone dots) such as a halftone image, there is a problem that the compression efficiency is deteriorated. This is because the discrete wavelet transform classifies the image into low-frequency components and high-frequency components, but many components remain in the high-frequency region in a specific halftone dot image, which is considered to be one of the causes that deteriorate the compression efficiency. (Ideally, the compression ratio improves when the high frequency component approaches uniform data close to 0).

  Therefore, as described above, the frequency characteristics of the tile image 51 are measured, the approximate pattern image 52 of the tile image 51 is generated using the frequency characteristics, and the difference image 53 between the tile image 51 and the approximate pattern image 52 is generated. By performing the discrete wavelet transform, the difference image 53 has less high-frequency component and low-frequency component data than the tile image 51, so that the compression rate can be improved.

<Second Embodiment>
In the first embodiment, the frequency characteristics of the tile image 51 are measured using Fourier transform or the like to generate the approximate pattern image 52, and then the image processing compresses the difference image 53 between the tile image 51 and the approximate pattern image 52. The apparatus 1 has been described. In the present embodiment, a plurality of approximate pattern images having different frequency characteristics are stored in advance in a storage unit, and an approximate pattern image having a high degree of coincidence with a tile image is selected from them, and a difference image is generated, and image compression is performed. An image processing apparatus that performs the above will be described.

  FIG. 7 is a block diagram showing an electrical configuration of the image processing apparatus 2 in the present embodiment. In FIG. 7, the same elements as those in the first embodiment are denoted by the same reference numerals, and description thereof is omitted. The storage unit 21 of the image processing unit 17 stores a compression program 41 and a decoding program 32 and further functions as a pattern image storage unit 42.

  The pattern image storage unit 42 stores a plurality of pattern images having different frequency characteristics. For example, the approximate pattern image 52 in FIG. 4 is an example, and a plurality of dot patterns having different numbers of lines are stored. In addition, it is desirable to investigate in advance the frequency characteristics of a pattern whose compression ratio is particularly bad and to store many pattern images that can be applied to an input image having the frequency characteristics.

  The pattern image storage unit 42 stores a plurality of pattern images having different frequency characteristics as table data, and the image processing unit 17 reads out the pattern image from this table as appropriate to generate a difference image. Then, the image processing unit 17 adds and stores the file name, pattern number, and the like of the pattern image that generated the difference image with the smallest data amount to the compressed difference image. In addition to storing a plurality of pattern images as table data, the pattern image storage unit 42 may have a basic expression that approximates a frequency component, and may store coefficients thereof.

  FIG. 8 is a flowchart showing the flow of compression processing performed when the image processing unit 17 reads and executes the compression program 41. First, the image processing unit 17 performs component conversion of the color system of the input image data, for example, from the RGB system to the YCbCr system (step S31), and a plurality of image data converted from the color system are displayed as shown in FIG. Is divided into tiles (step S32).

  Next, the image processing unit 17 generates a difference image between the plurality of pattern images stored in the pattern image storage unit 33 and the tile image (step S33), and the data amount is the smallest among the generated difference images. A difference image is selected (step S34). Here, the image processing unit 17 may simply select a difference image with a small amount of data, or may select it in consideration of a decrease in pixel values near the boundary of the tile image. The image processing unit 17 sets the pattern image that generated the selected difference image as an approximate pattern image, and stores it in the RAM 22 in association with the image information of the difference image (step S35).

  Subsequently, the image processing unit 17 performs discrete wavelet transform on the difference image selected in step S34 and outputs coefficient data (step S36), and performs linear quantization on the coefficient data subjected to the discrete wavelet transform (step S36). Step S37). Further, the image processing unit 17 performs entropy coding on the quantized quantized data and ends the process (step S38). Decoding is the same as in the first embodiment, and a description thereof will be omitted.

  As described above, a plurality of pattern images having different frequency characteristics are stored in advance in the pattern image storage unit 42, and the difference image having the smallest data amount among the difference images between the pattern images and the tile images is discretely obtained. By performing the wavelet transform, the difference image has less high-frequency component and lower-frequency component data than the tile image, so that the compression rate can be improved.

1 is a block diagram showing an electrical configuration of an image processing apparatus according to a first embodiment. The flowchart which shows the flow of the compression process in 1st Embodiment. The figure for demonstrating the process of JPEG2000. The figure for demonstrating the production | generation of the difference image in 1st Embodiment. The figure for demonstrating the process of JPEG2000. The flowchart which showed the flow of the decoding process in 1st Embodiment. The block diagram which shows the electrical constitution of the image processing apparatus in 2nd Embodiment. The flowchart which shows the flow of the compression process in 2nd Embodiment.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1, 2 Image processing apparatus 11 Control part 12 Storage part 13 Input operation part 14 Display part 15 I / F part 16 Network I / F part 17 Image processing part (Difference image acquisition means, wavelet transformation means, quantization means, frequency component Detection means, pattern image generation means, difference image generation means, inverse quantization means, inverse wavelet transform means, restoration means)
21 storage units 31, 41 compression program (image processing program)
32 Decoding program 43 Pattern image storage unit (pattern image storage means)
22 RAM (second storage means)

Claims (5)

Difference image acquisition means for acquiring a difference image between an input image and an approximate pattern image of the input image;
An approximate pattern image storage means for storing the approximate pattern image and the image information of the difference image in association with each other;
Wavelet transform means for applying wavelet transform to the difference image and outputting coefficient data;
Quantization means for quantizing the coefficient data and outputting quantized data;
Encoding means for encoding the quantized data and outputting the encoded data;
Frequency component detection means for detecting a frequency component of the input image;
Pattern image generation means for generating the approximate pattern image based on the detected frequency component;
Wherein the difference image acquiring unit images processing device for acquiring a difference image between the approximate pattern image the generated and the input image. Difference image acquisition means for acquiring a difference image between an input image and an approximate pattern image of the input image;
An approximate pattern image storage means for storing the approximate pattern image and the image information of the difference image in association with each other;
Wavelet transform means for applying wavelet transform to the difference image and outputting coefficient data;
Quantization means for quantizing the coefficient data and outputting quantized data;
Encoding means for encoding the quantized data and outputting the encoded data;
Pattern image storage means for storing a plurality of pattern images showing different frequency components;
Difference image generation means for generating a difference image between each of the stored pattern images and the input image;
Wherein the difference image obtaining means, the most small amount of data images processing device for acquiring by selecting the difference image of the generated difference image. De-encoding means for de-encoding the encoded data and outputting quantized data;
Inverse quantization means for inversely quantizing the quantized data and outputting coefficient data;
Inverse wavelet transform means for performing inverse wavelet transform on the coefficient data and outputting a difference image;
It reads the approximate pattern image the image information of the difference image from the approximate pattern image storing means is stored in association, and restoring means for restoring the differential image to the input image, further claim 1 or with a 2. The image processing apparatus according to 2. Computer
Difference image acquisition means for acquiring a difference image between an input image and an approximate pattern image of the input image;
Wavelet transform means for applying wavelet transform to the difference image and outputting coefficient data;
Quantization means for quantizing the coefficient data and outputting quantized data,
Encoding means for encoding the quantized data and outputting the encoded data;
A frequency component detecting means for detecting a frequency component of the input image;
Pattern image generation means for generating the approximate pattern image based on the detected frequency component;
To function as, the the difference image acquiring unit images processing program for obtaining a difference image between the approximate pattern image the generated and the input image. Computer
Difference image acquisition means for acquiring a difference image between an input image and an approximate pattern image of the input image;
An approximate pattern image storage means for storing the approximate pattern image and the image information of the difference image in association with each other;
Wavelet transform means for applying wavelet transform to the difference image and outputting coefficient data;
Quantization means for quantizing the coefficient data and outputting quantized data,
Encoding means for encoding the quantized data and outputting the encoded data;
Pattern image storage means for storing a plurality of pattern images showing different frequency components;
Difference image generation means for generating a difference image between each of the stored pattern images and the input image;
To function as the difference image acquiring unit, images processing program to select and retrieve small difference image the least data amount among the generated difference image.

Images