JP5244841B2 - Image processing apparatus, image processing method, and image processing program - Google Patents

Description

  The present invention relates to an image processing apparatus, an image processing method, and an image processing program, and more particularly to a technique for compressing image data.

  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 discrete wavelet transform, quantization, and entropy encoding for each tile. In addition, as described in Japanese Patent Application Laid-Open No. 2004-228620, a method for obtaining coefficients at each stage of discrete wavelet transform and creating version information for enhancing encoding is proposed as the generation of such encoded data. ing.

Japanese Patent Laid-Open No. 2007-166013

  The conventional general JPEG 2000 image compression method has improved compression efficiency compared to the JPEG method, but further improvement in compression efficiency is desired. Further, with respect to management of plate information by calculating coefficients in the encoded data generation method disclosed in Patent Document 1, further efficiency in encoding is required.

  The present invention has been made to solve the above problem, and an object of the present invention is to further improve the efficiency of encoding when a discrete wavelet transform is used in a compression algorithm such as JPEG2000.

The invention according to claim 1 of the present invention includes a color system conversion unit that converts image data into a color system composed of a luminance component and a color difference component;
A dividing unit that divides the image data converted by the color system conversion unit into a plurality of tiles;
A filter having a filter size used in each step of the discrete wavelet transform performed on the image data of each tile is determined and stored in advance , and the reference range fits in the pixel group size of the image data in each step for each step. A filter of a filter size, which is switched to a filter having a predetermined filter size in each step, and has a smaller reference range than the filter of the switched filter size and the predetermined filter. A discrete wavelet transform unit that performs discrete wavelet transform at each stage using another filter ;
A quantization unit that quantizes image data that has undergone discrete wavelet transform by the discrete wavelet transform unit;
An image processing apparatus comprising: an encoding unit that encodes the image data quantized by the quantization unit and outputs compressed data.

The invention according to claim 6 is a color system conversion step of converting image data into a color system consisting of a luminance component and a color difference component;
A division step of dividing the image data converted by the color system conversion step into a plurality of tiles;
A filter having a filter size used in each step of the discrete wavelet transform performed on the image data of each tile is determined and stored in advance , and the reference range fits in the pixel group size of the image data in each step for each step. A filter of a filter size, which is switched to a filter having a predetermined filter size in each step, and has a smaller reference range than the filter of the switched filter size and the predetermined filter. A discrete wavelet transform step for performing discrete wavelet transform at each stage using another filter ;
A quantization step for quantizing the image data that has undergone the discrete wavelet transformation by the discrete wavelet transformation step;
And an encoding step of encoding the image data quantized in the quantization step and outputting compressed data.

The invention according to claim 7 provides a computer,
A color system conversion unit that converts image data into a color system consisting of a luminance component and a color difference component;
A dividing unit that divides the image data converted by the color system conversion unit into a plurality of tiles;
A filter having a filter size used in each step of the discrete wavelet transform performed on the image data of each tile is determined and stored in advance , and the reference range fits in the pixel group size of the image data in each step for each step. A filter of a filter size, which is switched to a filter having a predetermined filter size in each step, and has a smaller reference range than the filter of the switched filter size and the predetermined filter. A discrete wavelet transform unit that performs discrete wavelet transform at each stage using another filter ;
A quantization unit that quantizes image data that has undergone discrete wavelet transform by the discrete wavelet transform unit;
It is an image processing program that functions as an encoding unit that encodes image data quantized by the quantization unit and outputs compressed data.

  In these inventions, at the time of discrete wavelet transform, for each stage, a filter size is selected according to the size of the pixel group size that can be referred to in the image data at each stage, and a filter having the selected filter size is used. In order to perform the discrete wavelet transform, for example, it is possible to perform the discrete wavelet transform by switching to a filter having a sufficient reference range with the pixel group size of the image data at each stage of the discrete wavelet transform. This reduces the situation where the number of discrete wavelet transform stages is limited by the size of the filter used for the discrete wavelet transform, and allows the number of discrete wavelet transform processing stages to be increased, improving the separation performance between high frequency and low frequency. Data continuity can be improved. Therefore, according to the present invention, it is possible to improve the encoding efficiency when using discrete wavelet transform in a compression algorithm such as JPEG2000.

  According to the present invention, the discrete wavelet transform unit stores in advance the filter used in each stage and performs the discrete wavelet transform using the predetermined filter in each stage. The number of stages of discrete wavelet transform is limited by the size of the filter used for discrete wavelet transform by storing a filter having a reference range sufficient for the pixel group size that can be referred to in each stage as each filter used in each stage. Can reduce the situation.

The invention of claim 2 of the invention is an image processing apparatus according to claim 1, wherein the discrete wavelet transform unit, to a predetermined phase from the first stage, the predetermined Discrete wavelet transform is performed using a filter of a filter size, and from the predetermined stage, the discrete wavelet transform is performed by changing to a different filter having a smaller reference range than the predetermined filter. Is.

  In this invention, the discrete wavelet transform unit uses a filter having a predetermined filter size for the discrete wavelet transform from the first stage to the predetermined stage without switching the filter at each stage. Since the reference range is changed to another filter of a smaller size from the determined stage, the sign bit for enabling the filter switching in each stage can be made unnecessary, and the predetermined stage is If the reference range of the filter used so far is set at a stage where the pixel group size that can be referred to for image data is insufficient, the number of stages of discrete wavelet transform can be reduced by the filter size. .

The invention according to claim 3 of the present invention is the image processing device according to claim 1, wherein the discrete wavelet transform unit refers to the predetermined filter used in the previous discrete wavelet transform. range, when you run out in the pixel group that fits the reference range of the filter for the image data of the current stage, the filter used for the discrete wavelet transform stage, from the filter used in the previous step, the predetermined Unlike the filter, the reference range is changed to another filter having a smaller filter size.

  In the present invention, the discrete wavelet transform unit is configured to detect the current pixel group size that can be referred to for the current stage image data when the reference range of the pixel that the filter used in the previous discrete wavelet transform needs to refer to is insufficient. Since the filter used for the discrete wavelet transform of the stage is changed from the filter used in the previous stage to another filter having a smaller reference range, the situation where the number of stages of the discrete wavelet transform is limited by the filter size can be reduced. .

The invention described in Claim 4 of the invention is an image processing apparatus according to claim 1, wherein the discrete wavelet transform unit, the at each stage, the predetermined filter of the filter size, and the A filter having a smaller coding size in each band in the image data after the discrete wavelet transform is performed by using both of a filter having a smaller reference size and a different filter size than a predetermined filter. Is a filter at this stage.

  In this invention, the discrete wavelet transform unit performs discrete wavelet transform at each stage using both a filter having a predetermined filter size and another filter having a smaller reference range than the filter. The filter with the smaller encoding size of each band of image data after wavelet transform is used as the filter at that stage, and at each stage, discrete wavelet transform is performed with the filter with the better separation performance between high frequency and low frequency. Since it is employed, the encoding efficiency can be further improved.

The invention according to claim 5 is the image processing apparatus according to claim 4 , wherein the discrete wavelet transform unit encodes each band in the image data after the discrete wavelet transform at each stage. Is determined based on the coding size for the band of the highest frequency component.

  In this invention, the discrete wavelet transform unit performs the determination of the coding size of each band in the image data after the discrete wavelet transform at each stage based on the coding size for the band of the highest frequency component, The processing time is shortened while maintaining the accuracy in determining the coding size.

  According to the present invention, it is possible to improve the encoding efficiency when using discrete wavelet transform in a compression algorithm such as JPEG2000.

It is a block diagram which shows the electric constitution of the image processing apparatus in this Embodiment. It is a figure which shows schematic structure of an image compression part. It is a figure which shows tile division. It is a schematic diagram shown about a discrete wavelet transform. It is the flowchart which showed 1st Embodiment of the image compression process which an image compression part performs according to an image compression program. (A)-(d) is a figure which shows the relationship between the pixel group size which can be referred by each tile size, and the reference range of the pixel which the filter used for a discrete wavelet transform requires. It is the flowchart which showed 2nd Embodiment of the image compression process which an image compression part performs according to an image compression program. It is the flowchart which showed 3rd Embodiment of the image compression process which an image compression part performs according to an image compression program. 14 is a flowchart illustrating a fourth embodiment of an image compression process performed by an image compression unit according to an image compression program 121;

  Embodiments of an image processing apparatus, an image processing method, 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, the information processing apparatus as the image processing apparatus 1 is 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. Also good.

  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 compression unit 17, a RAM 18, and the like. It is prepared for. The control unit 11 is configured by a CPU (Central Processing Unit) or the like, and reads out an image processing program stored in the storage unit 12 in accordance with an input instruction signal or the like, and executes processing. The image processing apparatus 1 is controlled in an integrated manner by outputting an instruction signal, data transfer, and the like.

  The storage unit 12 stores programs, data, and the like for realizing various functions provided in the image processing apparatus 1. In the present embodiment, the storage unit 12 stores an image compression program (an example of an image processing program referred to in the claims) 121. The image compression program 121 is a program for compressing the image data input by the I / F unit 15 or the network I / F unit 16 using the JPEG2000 method to generate encoded data. For example, the control unit 11 including a CPU or the like functions as the image compression unit 17 by operating according to the image compression program 121. The image compression program 121 is stored in the storage unit 12 by reading from a CD-ROM or DVD, or by downloading from the server on the Internet by the I / F unit 15 or the network I / F unit 16.

  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 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 compression unit 17 executes the image compression program 121 stored in the storage unit 12 to compress the image data input via the I / F unit 15 or the network I / F unit 16 using the JPEG 2000 method. Generate encoded data. A RAM (Random Access Memory) 18 temporarily stores a control program to be executed by the control unit 11, data for executing processing, and the like. In addition, the image compression unit 17 temporarily stores encoded data, compression rate adjustment data, and the like generated when processing is performed according to the image compression program 121.

  FIG. 2 is a diagram illustrating a schematic configuration of the image compression unit 17. FIG. 3 is a diagram showing tile division. FIG. 4 is a schematic diagram showing the discrete wavelet transform.

  The image compression unit 17 includes a color system conversion unit 171, a tile division unit (division unit) 172, a discrete wavelet conversion unit (discrete wavelet conversion unit) 173, a quantization unit 174, an encoding unit 175, It functions as the stream unit 176.

  First, the color system conversion unit 171 converts the color system of the input image data from, for example, the RGB system to the YCbCr system.

  Subsequently, the tile dividing unit 172 divides the image data whose color system has been converted into a plurality of tiles as shown in FIG. Here, the tile dividing unit 172 divides the tiles arranged in the main scanning direction as one band. The processing from the discrete wavelet transform unit 173 to the code stream unit 176 is performed for each band. Referring to FIG. 3, tiles A to E, tiles F to K, tiles L to Q, and tiles R to V are each divided into one band, and the processing after the discrete wavelet transform unit 173 is performed for each band. Therefore, it is assumed that all the processes described below are performed on a band basis.

  Returning to FIG. 2, the discrete wavelet transform unit 173 performs discrete wavelet transform on each tile of one band and outputs coefficient data. FIG. 4 is a schematic diagram showing the discrete wavelet transform. The discrete wavelet transform unit 173 performs discrete wavelet transform on the tile image (0LL) of FIG. 4A and decomposes it into subbands 1LL, 1HL, 1LH, and 1HH as shown in FIG. 4B. 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. Although FIG. 4 illustrates two-level conversion (two-stage conversion), the number of discrete wavelet transforms is not particularly limited.

  Then, the quantization unit 174 performs linear quantization on the coefficient data subjected to discrete wavelet transform. Further, the encoding unit 175 performs entropy encoding on the quantized data quantized at each compression rate. This entropy coding is performed by an algorithm called EBCOT (Embedded Block Coding with Optimized Truncation).

  After the entropy encoding is completed, the code stream unit 176 forms a code stream for writing data to the file. Since this code stream processing is well known from each document, the description thereof is omitted.

  FIG. 5 is a flowchart showing a first embodiment of an image compression process executed by the image compression unit 17 according to the image compression program 121. FIGS. 6A to 6D are diagrams illustrating the relationship between the pixel group size that can be referred to in each tile size and the reference range of pixels that are required by the filter used for the discrete wavelet transform. Note that the image processing shown in each flowchart for each of the following embodiments is an example of an image processing method in the claims.

  First, the color system conversion unit 171 and the tile division unit 172 perform color space conversion and tile division of image data, and divide the divided tiles by band (S1). Subsequently, the discrete wavelet transform unit 173 selects a band to be subjected to the image compression process (S2). Further, the discrete wavelet transform unit 173 selects a tile to be subjected to the discrete wavelet transform from the selected band (S3).

  Thereafter, the discrete wavelet transform unit 173 selects a filter to be used for the discrete wavelet transform in the first stage (S4), and performs a discrete wavelet transform on the image data of the tile using the selected filter (S5).

  For example, a case where the size of the divided tile is 128 × 128 (the size of an image input as a target of discrete wavelet transform) will be described. The discrete wavelet transform unit 173 stores a 9 (low frequency) / 7 (high frequency) size filter as a filter used for the discrete wavelet transform in the first stage (input image size 128 × 128). 9 (low frequency) / 7 (high frequency) size filters are stored as filters used in two stages (input image size 64 × 64), and used in the third step (input image size 32 × 32). A 9 (low frequency) / 7 (high frequency) size filter is stored as a filter, and a 9 (low frequency) / 7 (high frequency) filter is used in the fourth stage (input image size 16 × 16). A filter of size 9 (low frequency) / 7 (high frequency) is stored as a filter used in the fifth stage (input image size 8 × 8), and the sixth stage Filter used for (input image size 4x4) As a result, a filter having a size of 5 (low frequency) / 3 (high frequency) is stored.

  That is, the discrete wavelet transform unit 173 determines the pixel group size that can be referred to with the initial tile size and the reference range of the pixels required by the filter used for the discrete wavelet transform from FIGS. As shown in FIG. 4, the filter has a reference range having a size sufficient for a pixel group size that can be referred to in the current stage image data, and is stored so as to be used from the reference range having a larger reference range. The discrete wavelet transform unit 173 selects, for each stage, a filter having a size stored in advance in association with each stage, and uses it for the discrete wavelet transform at that stage.

  As described above, the discrete wavelet transform unit 173 selects a filter having a predetermined size for each stage at each stage, and performs discrete processing using the selected filter until reaching a predetermined final stage. The wavelet transform is repeated (NO in S6). That is, the discrete wavelet transform unit 173 stores in advance the final stage set as a stage where the pixel group size is not sufficient for the reference range of the filter, and determines whether or not the stage has been reached in S6. . If the discrete wavelet transform unit 173 determines that the stage of the discrete wavelet transform has reached the final stage (YES in S6), the quantization unit 174 quantizes the image data after the discrete wavelet transform (S7). The encoding unit 175 encodes (S8).

  Then, the discrete wavelet transform unit 173 determines whether or not an unprocessed tile remains for the band that is the target of image processing at this time (S9), and if an unprocessed tile remains (YES in S9). ), The processing from S3 to S9 is repeated for the next tile.

  On the other hand, if no unprocessed tile remains (NO in S9), it is determined whether or not a band that has not been subjected to discrete wavelet transform remains among the divided bands (S10). (YES in S10), the process returns to S2, and the processes of S3 to S9 are performed for the remaining bands.

  FIG. 7 is a flowchart showing a second embodiment of an image compression process executed by the image compression unit 17 in accordance with the image compression program 121. Note that the description of the same processing as in the first embodiment is omitted.

  In the second embodiment, the discrete wavelet transform unit 173 performs discrete wavelet transform using a filter having a predetermined filter size from the first stage to a predetermined stage, and from the predetermined stage. From the subsequent stage, the discrete wavelet transform is performed by changing to another predetermined filter having a smaller reference range (S25, S26, S27).

  That is, in the first embodiment, the discrete wavelet transform unit 173 selects a filter to be used at each stage for each stage, but in this second embodiment, the discrete wavelet transform unit 173 performs discrete wavelet transform. It is determined whether or not the stage has reached a predetermined stage (S26), and the filter of the size selected at the time of the first stage is used until the predetermined stage is reached (YES in S26) ( NO in S26, S25), from the predetermined stage, another predetermined filter having a smaller reference range is selected (S27), and discrete wavelet transform is performed using the changed filter. (S28).

  In the second embodiment, the discrete wavelet transform unit 173 determines, as the predetermined stage, a stage in which the pixel group size that can be referred to in the image data after the discrete wavelet transform is insufficient in the reference range of the filter. It is stored in advance as a given stage, and it is determined in S26 whether or not the stage has been reached. For example, in the example shown in FIGS. 6A to 6D, from the stage of an input image size of 8 × 8 that can be processed by a filter having a size of 9 (low frequency) / 7 (high frequency), 9 (Low frequency) / 7 (high frequency) size processing is impossible, and 5 (low frequency) / 3 (high frequency) size processing is possible with an input image size of 4x4 The discrete wavelet transform unit 173 stores the step of the input image size 4 × 4 in the case of the above as the predetermined step.

  FIG. 8 is a flowchart showing a third embodiment of the image compression processing executed by the image compression unit 17 in accordance with the image compression program 121. Note that description of the same processing as in the first embodiment or the second embodiment is omitted.

  In the third embodiment, the discrete wavelet transform unit 173 performs discrete wavelet transform using a filter having a predetermined filter size from the first stage, and after the discrete wavelet transform at each stage, the discrete wavelet transform unit 173 However, it is determined whether or not the reference range of the filter used in the discrete wavelet transform in the previous stage is sufficient for the pixel group size that can be referred to for the current stage image data (S46), and used in the discrete wavelet transform in the previous stage. When the reference range of the filter is insufficient for the pixel group size that can be referred to in the current stage image data (NO in S46), the filter used for the discrete wavelet transform of the current stage is changed from the filter used in the previous stage. Change to another predefined filter with a smaller reference range ( 47). From this stage, the discrete wavelet transform unit 173 performs discrete wavelet transform using the changed filter (S48).

  That is, in the third embodiment, the discrete wavelet transform unit 173 does not store the filter used at each stage in advance, and changes the filter indicated as S46 every time the discrete wavelet transform at each stage is completed. In the example shown in FIGS. 6A to 6D, an input image size of 8 × 8 that can be processed with a filter of size 9 (low frequency) / 7 (high frequency) is used. Input image size 4 that cannot be processed with a 9 (low frequency) / 7 (high frequency) size filter and that can be processed with a 5 (low frequency) / 3 (high frequency) size filter. Judge whether or not it is in the x4 stage.

  FIG. 9 is a flowchart showing a fourth embodiment of the image compression processing executed by the image compression unit 17 according to the image compression program 121. Note that the description of the same processes as those in the first to third embodiments is omitted.

  In the fourth embodiment, the discrete wavelet transform unit 173 uses both a filter having a predetermined filter size and a filter having another predetermined filter size whose reference range is smaller than that of the filter at each stage. The discrete wavelet transform is performed (S64, S65). After the discrete wavelet transform using the two filters, the discrete wavelet transform unit 173 has a small encoding size for the image data of each band (LL, HL, LH, HH) for the image data after the discrete wavelet transform. The filter that has become the discriminant is discriminated (S66), and the image data that has been subjected to the discrete wavelet transform by the filter is used as the image data at this stage. That is, the filter with the smaller coding size of each band after the discrete wavelet transform is adopted as the filter at this stage (S67).

  For example, the discrete wavelet transform unit 173 performs a discrete wavelet transform using both a 9 (low frequency) / 7 (high frequency) size filter and a 5 (low frequency) / 3 (high frequency) size filter, Among the two pieces of image data after the discrete wavelet transform, a filter having a smaller encoding size is determined for the image data of each band (LL, HL, LH, HH).

  In S64, when the discrete wavelet transform unit 173 selects a filter to be used at each stage, the pixel group size of the image data at each stage is set to the predetermined filter size (for example, 9 (low frequency) / 7). It is determined whether the input image size can be processed with a filter of (high frequency)), and processing with the filter of the predetermined filter size is impossible, and the reference range is larger than the filter. Is an input image size that can be processed by a filter having a different filter size (for example, a size of 5 (low frequency) / 3 (high frequency)), which is smaller, in S65, the reference range. The discrete wavelet transform is performed by using a filter having a predetermined filter size smaller than. In this case, the processing of S66 and S67 is not performed, and the processing of S68 is performed.

  In the processes of S64 to S68, the stage of the discrete wavelet transform unit is changed to the final stage (the pixel group size of the image data after the discrete wavelet transform is changed to a reference range of a filter of another predetermined size in which the reference range is further smaller. The process is continued until it is reached (S68).

  In S66, the discrete wavelet transform unit 173 may determine the coding size of each band after the discrete wavelet transform at each stage based on the coding size of the highest band (HH) component. . In this case, the processing time can be shortened while maintaining accuracy in determining the coding size of each band in the image data after the discrete wavelet transform.

  The present invention is not limited to the configuration of the above embodiment, and various modifications can be made. For example, each of the above embodiments can be applied to the case where either reversible transformation or irreversible transformation is performed in discrete wavelet transformation. In particular, in the past, in the case of irreversible transform compression, the filter used for discrete wavelet transform was fixed (for example, the filter size of 9 (low frequency) / 7 (high frequency)). When the pixel group size of the image data after the discrete wavelet transform is insufficient for the reference range of the fixed filter, another filter (for example, 5 (low frequency ) / 3 (high frequency) size filter), it is possible to increase the number of discrete wavelet transforms (number of stages), so pixels in the part containing high frequency components such as HL, LH, HH The continuity of values increases, and the compression efficiency can be increased.

  The configuration and processing according to the embodiment illustrated in FIGS. 1 to 9 are merely examples of the configuration and processing of the image processing apparatus, the image processing method, and the image processing program according to the present invention. The configuration and processing of the image processing apparatus, the image processing method, and the image processing program are not limited to the contents described above.

DESCRIPTION OF SYMBOLS 1 Image processing apparatus 11 Control part 12 Storage part 13 Input operation part 14 Display part 17 Image compression part 171 Color system conversion part 172 Tile division part 173 Discrete wavelet transformation part 174 Quantization part 175 Encoding part 176 Code stream part

Claims (7)

A color system conversion unit that converts image data into a color system consisting of a luminance component and a color difference component;
A dividing unit that divides the image data converted by the color system conversion unit into a plurality of tiles;
A filter having a filter size used in each step of the discrete wavelet transform performed on the image data of each tile is determined and stored in advance , and the reference range fits in the pixel group size of the image data in each step for each step. A filter of a filter size, which is switched to a filter having a predetermined filter size in each step, and has a smaller reference range than the filter of the switched filter size and the predetermined filter. A discrete wavelet transform unit that performs discrete wavelet transform at each stage using another filter ;
A quantization unit that quantizes image data that has undergone discrete wavelet transform by the discrete wavelet transform unit;
An image processing apparatus comprising: an encoding unit that encodes the image data quantized by the quantization unit and outputs compressed data. The discrete wavelet transform unit, to a predetermined phase from the first stage performs a discrete wavelet transform using the filter of a filter having a predetermined size, from the predetermined phase, the predetermined The image processing apparatus according to claim 1 , wherein the discrete wavelet transform is performed by changing to another filter having a filter size with a smaller reference range unlike the filter . The discrete wavelet transform unit, when the reference range of the predetermined filter used in the discrete wavelet transform of the previous stage is insufficient in the pixel group size that fits the reference range of the filter for the current stage image data, 2. The image processing according to claim 1, wherein the filter used for the discrete wavelet transform at the current stage is changed from the filter used in the previous stage to another filter having a smaller reference range than the predetermined filter. apparatus. The discrete wavelet transform unit, the in each stage, a different filter, discrete wavelet using both of the predetermined filter of the filter size, and the previously reference range differs from the filter defined is smaller filter size The image processing apparatus according to claim 1, wherein a filter having a smaller encoding size of each band in the image data after the discrete wavelet transform is used as a filter at the stage. 5. The discrete wavelet transform unit according to claim 4 , wherein the discrete wavelet transform unit determines the coding size of each band in the image data after the discrete wavelet transform in each stage based on the coding size for the band of the highest frequency component. Image processing device. A color system conversion step for converting image data into a color system consisting of a luminance component and a color difference component;
A division step of dividing the image data converted by the color system conversion step into a plurality of tiles;
A filter having a filter size used in each step of the discrete wavelet transform performed on the image data of each tile is determined and stored in advance , and the reference range fits in the pixel group size of the image data in each step for each step. A filter of a filter size, which is switched to a filter having a predetermined filter size in each step, and has a smaller reference range than the filter of the switched filter size and the predetermined filter. A discrete wavelet transform step for performing discrete wavelet transform at each stage using another filter ;
A quantization step for quantizing the image data that has undergone the discrete wavelet transformation by the discrete wavelet transformation step;
An image processing method comprising: an encoding step of encoding the image data quantized in the quantization step and outputting compressed data. Computer
A color system conversion unit that converts image data into a color system consisting of a luminance component and a color difference component;
A dividing unit that divides the image data converted by the color system conversion unit into a plurality of tiles;
A filter having a filter size used in each step of the discrete wavelet transform performed on the image data of each tile is determined and stored in advance , and the reference range fits in the pixel group size of the image data in each step for each step. A filter of a filter size, which is switched to a filter having a predetermined filter size in each step, and has a smaller reference range than the filter of the switched filter size and the predetermined filter. A discrete wavelet transform unit that performs discrete wavelet transform at each stage using another filter ;
A quantization unit that quantizes image data that has undergone discrete wavelet transform by the discrete wavelet transform unit;
An image processing program that functions as an encoding unit that encodes image data quantized by the quantization unit and outputs compressed data.

Images