JP4874941B2 - Image processing apparatus and image processing method - Google Patents

Description

  The present invention relates to the field of image processing, and more particularly to an image encoding technique and an image encoded data decoding technique.

  Due to the improvement in accuracy of digital cameras and scanners, the demand for handling high-definition still images has increased, and the functions required for image compression / expansion techniques to facilitate the handling of high-definition still images have diversified.

JPEG (Joint Photographic) is an image compression / decompression algorithm for high-definition still images.
Experts Group) is currently most frequently used, but in recent years, image compression / decompression algorithms using discrete wavelet transform (DWT) as frequency transform have increased, and a representative example is JPEG2000.

  One of the major features of JPEG2000 is resolution scalability. Resolution scalability works preferentially when it is desired to view a target image after converting it to a desired image size. The image size of the high-definition still image to be displayed is determined by the accuracy (dot pitch) and size when reading a signal from a digital camera or scanner. Once the data is adapted to a specific still image format, it is constant. The display size (resolution) when browsing this image varies.

  When an image encoded by a method that does not have resolution scalability such as JPEG is to be displayed in a desired size, generally, the entire encoded data is subjected to a decoding process and expanded into pixel data (bitmap). Therefore, it is necessary to perform enlargement / reduction processing. On the other hand, when displaying an image encoded by JPEG2000 in a reduced size, pixel data (bitmap) of the reduced image can be generated by performing decoding processing (reduction decoding) of only a part of the encoded data. The memory required for the decoding process can be reduced and the decoding process time can be shortened. For example, Patent Document 1 is a prior art document relating to such reduced decoding.

JP 2003-338932 A

  When a binary image is lossless-encoded with JPEG2000 and the encoded data is reduced and decoded, there is a problem that thin lines in the original image disappear or are interrupted. This problem is due to the discrete wavelet transform during the JPEG2000 encoding process. That is, when calculating the wavelet coefficient, a rounding process is performed to make the coefficient an integer, and then every other pixel is down-sampled. Therefore, there is no problem in a region where continuous pixel values continue, but black thin lines appear on the white ground. This is because the black pixel component may disappear in the conversion process because the continuous width of the pixel value is small when it exists.

  Further, even when the encoded data obtained by lossless encoding the multi-valued image with JPEG2000 is reduced and decoded, there is a problem that the reproducibility of the fine line is poor if it is not as much as in the case of the binary image. Due to the discrete wavelet transform.

  The present invention is intended to improve the above-mentioned problems, and more specifically, by using an encoding method having resolution scalability such as JPEG2000 that uses discrete wavelet transform in the encoding process, the fine line at the time of reduced decoding is reduced. An object of the present invention is to provide an image processing apparatus and method for generating encoded data with good reproducibility. Another object is to provide an image processing apparatus or method for restoring an image in which fine lines in the original image are well reproduced from encoded data generated by such an image processing apparatus or method. .

  In the case of JPEG2000 lossless encoding, for example, an image is divided into tiles, and one-dimensional lossless discrete wavelet transform is applied to each tile image in the horizontal and vertical directions. Normally, taking into account downsampling after the discrete wavelet transform, a lifting operation such as the following equation is used.

ここで、Xextは、タイル境界の外の画素を参照する際に不連続性を抑えて滑らかにするために１次元入力信号を拡張した後の信号である。式（１）により高域成分をダウンサンプリングしたものが求まり、これを用いて式（２）により低域成分をダウンサンプリングしたものが求まる。

Here, Xext is a signal obtained by extending the one-dimensional input signal in order to suppress discontinuity and smooth when referring to pixels outside the tile boundary. The one obtained by down-sampling the high-frequency component is obtained from Equation (1), and the one obtained by down-sampling the low-frequency component is obtained from Equation (2) using this.

  It should be noted here that, in the case of a binary image in which a thin line having a width of one pixel as shown in FIG. In the required low frequency component, the element “1” disappears.

  As can be seen from equation (2), the low frequency components are based on the even coordinate values of the input signal, and the high frequency components of the adjacent odd coordinate values obtained from equation (1) are both “1”. This is because a value different from the pixel value is used as a coefficient only in the case of That is, the low-frequency component output by the discrete wavelet transform for the binary image matches the pixel value of the even coordinate value at all points except the point where both of the adjacent high-frequency components are “1”. As a result, the thin line of 1 pixel width that existed in the odd coordinate value disappears from the low frequency component when the discrete wavelet transform is performed once, and conversely, the thin line of 1 pixel width that existed in the even coordinate value also saved in the low frequency component. Is done. The above is the same even when the discrete wavelet transform is applied two-dimensionally (in the horizontal and vertical directions).

  Here, consider a translation image obtained by translating the original image by, for example, one pixel in the horizontal and vertical directions. Pixels with even coordinate values on the original image correspond to pixels with odd coordinate values on the translation image. Therefore, when a discrete wavelet transform is performed in the horizontal and vertical directions with the original image as the component C0 and the translation image as the component C1, the thin line disappeared from the low frequency component (LL subband) of the component C0 is the low frequency component of the component C1. On the contrary, the thin line lost from the low frequency component of the component C1 is stored in the low frequency component of the component C0.

  Assuming that the discrete wavelet transform is recursively applied three times to the original image (component C0) and the translation image (component C1) obtained by translating the original image by one pixel in the horizontal and vertical directions, respectively. Each component is decomposed into subband coefficients as shown in FIG. Consider a case where encoded data having resolution scalability of two components is generated by entropy encoding such a transform coefficient and decoded.

  When the decoding resolution level is the highest resolution level (here, resolution level 3), the lossless original image is restored by decoding the code of the component C0. Therefore, the code of component C1 does not need to be decoded.

  In the case of reduced decoding at a resolution level 2 where the decoding resolution level is one level lower, if the codes of the components C0 and C1 are decoded, the 1-pixel-wide thin line in the original image is stored in one of the restored components C0 and C1. Has been. At resolution level 2, an adjacent point shifted by one pixel on the original image overlaps the same point on components C0 and C1. Therefore, by performing the overlapping process of the restored components C0 and C1, it is possible to generate an image in which a thin line having a width of one pixel on the original image is restored. When each component is a binary image, the superimposition process of the components C0 and C1 can be a process of calculating a logical sum of pixel values at the same position. When each component is a multi-valued image, the superimposing process of the components C0 and C1 can be a process that takes the maximum value among the pixel values at the same position.

  In the case of decoding at the highest resolution level 3, it is not preferable to decode the code of the component C1 and perform the overlapping process of the restored components C0 and C1. This is because the component C0 and the component C1 are shifted by one pixel, and therefore, when the superimposition process is performed, an image appears as a doubled original image.

  Thus, for component C1, the code at resolution level 3 is not useful for decoding. Therefore, at the time of encoding, the transform coefficient of the resolution level 3 of the component C1, that is, the 1HL, 1LH, and 1HH subband coefficients in FIG. 24 may be forced to be 0, which reduces the code amount. be able to.

  Up to this point, the description has been made assuming that only one translation image obtained by translating the original image by one pixel is generated. However, a plurality of translation images obtained by moving a different predetermined number of pixels are generated, and these parallel images and By applying the discrete wavelet transform using the original image as a component and entropy coding the transform coefficient, it is possible to generate encoded data having three or more components. When decoding this encoded data, for example, by selecting a necessary component according to the decoding resolution level, performing decoding processing on the code, and performing overlapping processing of the restored component, a thin line It is possible to generate an image with good reproducibility (see the embodiment described later for details).

  The present invention is to achieve the above object based on the above consideration, and the features thereof are as follows.

That is, the image processing apparatus according to the invention of claim 1
Image input means for inputting images,
Image generating means for generating one or a plurality of translation images obtained by translating the original image from an image input by the image input means (hereinafter referred to as an original image);
An encoding process for generating encoded data having resolution scalability consisting of a plurality of components by performing discrete wavelet transform for each component, and entropy encoding the transform coefficient, with the original image and the parallel moving image as components. Means.

According to a second aspect of the present invention, there is provided an image processing apparatus according to the first aspect of the present invention.
The encoding processing means sets a value of a conversion coefficient equal to or higher than a resolution level determined by a translation amount of the translation image with respect to the original image, to 0 among the transformation coefficients of the translation image.

According to a third aspect of the present invention, there is provided an image processing apparatus according to the first aspect of the present invention.
The encoding processing means is to describe information for selecting an effective component in accordance with a decoding resolution level when decoding the encoded data in the header of the encoded data.

According to a fourth aspect of the present invention, there is provided an image processing apparatus according to the first aspect of the present invention.
The encoding processing means describes in the header of the encoded data information for selecting a component superposition method according to a decoding resolution level when decoding the encoded data.

According to a fifth aspect of the present invention, there is provided an image processing apparatus according to any one of the first to fourth aspects,
Storage means for storing encoded data generated by the encoding processing means;
Means for communication with external devices,
Based on the code request from the external device received by the communication unit, the code that needs to be transmitted to the external device is extracted from one encoded data stored in the storage unit, and the extracted code is It is further characterized by further comprising request processing means for transmitting to the external device by the communication means.

An image processing apparatus according to a sixth aspect of the present invention provides:
An original image and one or a plurality of translated images obtained by translating the original image are used as components, respectively, and discrete wavelet transform is performed for each component, and the transform coefficient is generated by entropy encoding, and includes a plurality of components. Code input means for inputting encoded data having resolution scalability;
A decoding resolution level specifying means for specifying a decoding resolution level;
Component selecting means for selecting an effective component of the encoded data (hereinafter referred to as input encoded data) input by the encoding input means in accordance with the decoding resolution level specified by the decoding resolution level specifying means;
Decoding processing means for performing decoding processing up to the decoding resolution level specified by the decoding resolution level specifying means, with the effective component selected by the component selecting means as a decoding target for the input encoded data;
Component superimposing processing means for generating a restored image by performing superimposing processing of the component restored by the decoding processing of the decoding processing means.

An image processing apparatus according to a seventh aspect of the present invention provides:
A resolution composed of a plurality of components generated by performing discrete wavelet transform for each component and entropy encoding the transform coefficient for each component, using the original image and one or a plurality of translated images obtained by translating the original image as components. Code input means for inputting encoded data having scalability;
A decoding resolution level specifying means for specifying a decoding resolution level;
Component selecting means for selecting an effective component of the encoded data (hereinafter referred to as input encoded data) input by the encoding input means in accordance with the decoding resolution level specified by the decoding resolution level specifying means;
Decoding processing means for performing decoding processing up to the decoding resolution level specified by the decoding resolution level specifying means, with all components as decoding targets for the input encoded data;
Component overlay processing means for generating a restored image by performing overlay processing of valid components selected by the component selection means in the components restored by the decoding processing of the decoding processing means. .

A feature of the invention described in claim 8 is that, in the image processing apparatus according to claim 6 or 7, the header of the input encoded data corresponds to a decoding resolution level when the input encoded data is decoded. Information to select valid and valid components,
The component selecting means selects an effective component based on the information described in the header of the input encoded data and the decoding resolution level specified by the decoding resolution level specifying means.

An image processing apparatus according to the invention of claim 9 is provided.
A resolution composed of a plurality of components generated by performing discrete wavelet transform for each component and entropy encoding the transform coefficient for each component, using the original image and one or a plurality of translated images obtained by translating the original image as components. Code input means for inputting encoded data having scalability;
A decoding resolution level specifying means for specifying a decoding resolution level;
Decoding processing up to the decoding resolution level designated by the decoding resolution level designation unit is performed on all the components as decoding targets for the coded data input by the code input unit (hereinafter referred to as input coded data). Decryption processing means to perform,
Component superimposing method selecting means for selecting a superimposing method of components restored in the decoding process of the decoding processing means according to the decoding resolution level designated by the decoding resolution level designating means,
The decoding processing means includes means for generating a restored image by performing a superimposition process on the component restored by the decoding process by the superimposing method of the components selected by the component superposition method selecting means. And

According to a tenth aspect of the present invention, there is provided an image processing apparatus according to the ninth aspect of the present invention.
In the header of the input encoded data, information for selecting a component superposition method according to a decoding resolution level when the input encoded data is decoded is described.
The component superposition method selection means selects a component superposition method based on the information described in the header of the input encoded data and the decoding resolution level designated by the decoding resolution level designation means. It is in.

An image processing method according to the invention of claim 11 is provided.
An image input process for inputting an image,
An image generation step of generating one or a plurality of parallel movement images obtained by translating the original image from the image input by the image input step (hereinafter referred to as an original image);
An encoding process for generating encoded data having resolution scalability consisting of a plurality of components by performing discrete wavelet transform for each component, and entropy encoding the transform coefficient, with the original image and the parallel moving image as components. A process.

An image processing method according to the invention of claim 12 is provided.
An original image and one or a plurality of translated images obtained by translating the original image are used as components, respectively, and discrete wavelet transform is performed for each component, and the transform coefficient is generated by entropy encoding, and includes a plurality of components. A code input process for inputting encoded data having resolution scalability;
A decoding resolution level specifying step for specifying a decoding resolution level;
A component selection step of selecting an effective component of the encoded data (hereinafter referred to as input encoded data) input by the encoding input step according to the decoding resolution level specified by the decoding resolution level specifying step;
A decoding process step of performing a decoding process up to the decoding resolution level designated by the decoding resolution level designation step, with the effective component selected by the component selection step as a decoding target for the input encoded data;
And a component superimposing process step of generating a restored image by performing a superimposing process of the component restored by the decoding process of the decoding process step.

An image processing method according to the invention of claim 13 is provided.
An original image and one or a plurality of translated images obtained by translating the original image are used as components, respectively, and discrete wavelet transform is performed for each component, and the transform coefficient is generated by entropy encoding, and includes a plurality of components. A code input process for inputting encoded data having resolution scalability;
A decoding resolution level specifying step for specifying a decoding resolution level;
A component selection step of selecting an effective component of the encoded data (hereinafter referred to as input encoded data) input by the encoding input step according to the decoding resolution level specified by the decoding resolution level specifying step;
A decoding process step for performing decoding processing up to the decoding resolution level specified by the decoding resolution level specifying step, with all components as decoding targets, for the input encoded data;
A component overlay processing step of generating a restored image by performing overlay processing of the effective components selected by the component selection step in the components restored by the decoding processing of the decoding processing step. .

An image processing method according to the invention of claim 14 is provided.
An original image and one or a plurality of translated images obtained by translating the original image are used as components, respectively, and discrete wavelet transform is performed for each component, and the transform coefficient is generated by entropy encoding, and includes a plurality of components. A code input process for inputting encoded data having resolution scalability;
A decoding resolution level specifying step for specifying a decoding resolution level;
Decoding processing up to the decoding resolution level specified in the decoding resolution level specifying step is performed on all the components as decoding targets for the encoded data input in the code input step (hereinafter referred to as input encoded data). A decryption processing step to be performed;
A component superposition method selection step of selecting a superposition method of components restored in the decoding processing of the decoding processing step according to the decoding resolution level designated by the decoding resolution level designation step;
The decoding process step includes a step of generating a restored image by performing a superimposition process on the component restored by the decoding process by the superimposing method of the component selected by the component superposition method selection step. And

  According to the first to fifth and eleventh aspects of the present invention, it is possible to generate encoded data with good reproducibility of fine lines at the time of reduced decoding and necessary for image restoration of such encoded data. Only the code can be transmitted to the external device. According to the inventions described in claims 6 to 10 and 13 to 14, it is possible to obtain an effect such that an image in which fine lines in an original image are well reproduced can be restored even in the case of reduced decoding.

Embodiments of the present invention will be described below with reference to the drawings.
[First Embodiment]
FIG. 1 is a block diagram of an image processing apparatus according to the first embodiment of the present invention. The image processing apparatus 100 includes an image input unit 101, a parallel movement image generation unit 102, and an encoding processing unit 103.

  The image input unit 101 is a means for inputting an image to be processed. Specifically, for example, the image input unit 101 is a scanner that scans a document and inputs an image, or a local storage (not shown) or a storage on a network. It is an interface that reads more images. The translation image generation unit 102 generates one or a plurality of images (referred to as translation images) obtained by translating the original image from the image input by the image input unit 101 (referred to as an original image). Means. The encoding processing unit 103 captures the original image and the parallel moving image from the parallel moving image generation unit 102, and performs encoding processing according to the encoding algorithm of JPEG2000, using these images as components, respectively, so that a plurality of components can be obtained. And JPEG2000 format encoded data having resolution scalability. The generated encoded data is accumulated in the storage 104, for example.

  The image processing apparatus 100 can be realized by software using, for example, hardware resources of a computer having a general configuration including a CPU and a memory.

  The translation image generation unit 102 generates one or a plurality of translation images obtained by translating the original image by a predetermined amount. The maximum number of translation images to be generated is determined by the discrete wavelet transform in the encoding processing unit 103. It is equal to the number of times of application (decomposition number) minus 1. For example, when the discrete wavelet transform is applied five times, a maximum of four translation images can be generated. If such a maximum number of parallel images are generated, it is possible to generate encoded data that can be used for reduced decoding up to the lowest resolution level 1.

  The amount of movement of the translation image relative to the original image is determined in consideration of the effect of downsampling in the discrete wavelet transform. For example, when the discrete wavelet transform is applied five times, a maximum of four translation images are generated, and the amount of movement of each translation image relative to the original image is one pixel, two in the first translation image. It can be selected from 2 or 3 pixels for the translation image of the eye, 4 to 7 pixels for the third translation image, and 8 to 15 pixels for the fourth translation image.

  The generation of a translation image will be described in accordance with the image representation method in JPEG2000. Consider an original image in which the offset from the origin (0, 0) of the reference grid as schematically shown in FIG. 2 is (XOsiz, YOsiz) and the horizontal and vertical sizes are Xsiz, Ysiz. An image as shown in FIG. 3 is created by cutting out one pixel at the left end and one pixel at the upper end of the original image, and the (0, 0) point of this image is displayed on the reference grid as shown in FIG. (XOsiz, YOsiz) to generate a translation image in which the original image is moved by one pixel in the horizontal and vertical directions. Note that the horizontal and vertical sizes of this translation image are smaller by one pixel than the original image, so that one pixel is compensated by a method (mirroring) of folding neighboring pixels at the lower end and the right end. Also good.

  Similarly, FIG. 5 schematically shows a translation image obtained by moving the original image by 3 pixels in the horizontal and vertical directions. In this parallel moving image, the horizontal and vertical sizes are smaller by 3 pixels than the original image, so that 3 pixels may be supplemented by mirroring at the lower end and the right end.

  Next, the encoding processing unit 103 will be described. The processing in the encoding processing unit is partially different between the case where the original image and the parallel movement image are binary images and the case of a multi-value image (for example, 8 bits / pixel). The present invention is most effective when the original image and the translation image are binary images.

  FIG. 6 shows a block diagram of the encoding processing unit 103 when processing a binary image. The encoding processing unit 103 performs encoding processing according to the encoding algorithm of JPEG2000. As is clear from FIG. 6, since the input image is a binary image, DC level shift processing for halving the dynamic range is performed. Not performed. Also, component conversion processing assuming a color image is not performed.

  First, in the processing block 111, the reversible discrete wavelet transform is applied in the vertical direction and the horizontal direction for each component, with the original image and the translation image as components. If specified, tile division of each component is performed, and a reversible discrete wavelet transform is applied to each tile. Here, it is assumed that four translation images are generated as described above. Component C0, translation image obtained by moving the translation image by 1 pixel, component C0, translation image obtained by translation by 3 pixels, component C2, translation image obtained by translation by 4 to 7 pixels, component C3, 8 to 15 pixels The parallel movement image that has been moved by the distance is called a component C4.

  If the discrete wavelet transform is applied five times, each component is decomposed into subband coefficients as shown in FIG. Although it is possible to immediately perform entropy coding on this transform coefficient, in the example shown here, in order to reduce the code amount, the value of a transform coefficient that is not useful (invalid) in decoding of components C1 to C4 in the processing block 112 Is set to 0.

  FIG. 8 shows the number of resolution levels of invalid conversion coefficients of the components C1 to C4. That is, in the component C1 having a movement amount of 1 pixel, conversion coefficients (1HL, 1LH, 1HH subband coefficients in FIG. 7) counted from the highest resolution level (in this example, resolution level 5), that is, resolution level 5 are converted. Is set to 0 as an invalid conversion coefficient. In the component C2, conversion coefficients (1HL, 1LH, 1HH, 2HL, 2LH, and 2HH subband coefficients in FIG. 7) at two resolution levels, that is, resolution levels 5 and 4, are set to 0 as invalid conversion coefficients. In component C3, conversion coefficients of three resolution levels, that is, resolution levels 5, 4 and 3 (1HL, 1LH, 1HH,..., 3HL, 3LH, 3HH subband coefficients in FIG. 7) are set to 0 as invalid conversion coefficients. Is set. Similarly, in the component C4, the conversion coefficients (1HL, 1LH, 1HH,..., 4HL, 4LH, 4HH subband coefficients in FIG. 7) of resolution levels 5, 4, 3, and 2 are set to 0 as invalid conversion coefficients. Is set.

  FIG. 9 is a diagram for explaining the relationship between the decoded resolution level and the effective resolution level of the code of each component when decoding the encoded data having the number of resolution levels of 5.

  When the decoding resolution level is the highest resolution level 5, the same (lossless) image as the original image is restored by performing the decoding process up to the resolution level 5 with only the code of the component C0 as the decoding target. In this case, for example, when decoding is performed up to the code of the component C1, and the restored components C0 and C1 are superimposed, the component C1 is an image shifted by one pixel from the original image. This image becomes an image in which lines in the original image are doubled.

  When the decoding resolution level is resolution level 4, by performing decoding processing (reduction decoding) up to resolution level 4 using the codes of components C0 and C1 as decoding targets, and performing overlapping processing of the restored components C0 and C1, A reduced image whose resolution is one level lower than that of the original image is restored. At the resolution level 4, adjacent pixels shifted by one pixel on the original image are in the same position on the components C0 and C1, so there is no inconvenience that the lines of the image after the overlapping process are doubled. The thin line disappeared from the component C0 or C1 by the discrete wavelet transform is also restored.

  As can be seen from the above, the code of the resolution level 5 of the component C1 is not used for decoding. Therefore, in the component C1, the conversion coefficient at the resolution level 5 may be invalid and the value may be set to 0.

  Similarly, when the decoding resolution level is resolution level 3, the codes of components C0, C1, and C2 are subjected to reduction decoding to resolution level 3, and the restored components C0, C1, and C2 are superimposed. By doing so, it is possible to restore a reduced image that is two levels lower in resolution than the original image. Then, the thin line disappeared from the component C0, C1, or C2 by the discrete wavelet transform is also restored. The codes of the resolution levels 5 and 4 of the component C2 are not used for decoding. Therefore, the conversion coefficients of the resolution levels 5 and 4 of the component C2 may be invalidated, and the value may be set to 0.

  The components C3 and C4 can be described in the same manner. The conversion coefficient of the resolution level 5, 4, 3 of the component C3 is set to 0, and the conversion coefficient of the resolution level 5, 4, 4, 2 of the component C4 is 0. Set to

  When the highest restoration resolution level is set to resolution level 4, components C2, C3, and C4 are unnecessary. When decoding is not performed at a resolution level lower than the resolution level 3, the components C3 and C4 are unnecessary. If decoding is not performed at the resolution level 1, the component C4 is unnecessary.

  If the code amount reduction effect is not expected, the processing block 112 can be omitted, and such an aspect is also included in the present invention.

  Next, in the processing block 113, entropy coding by coefficient bit modeling and arithmetic coding is performed on the transform coefficient of each component processed in the processing block 112. Next, in the processing block 114, a JPEG2000 format having resolution scalability consisting of a plurality of components is generated by combining packets generated by entropy encoding to generate packets, arranging the packets in a predetermined order, and adding a header or the like. Encoded data (code stream) is generated. Note that the packet corresponding to the transform coefficient whose value is set to 0 in the processing block 112 is a 0-length packet, so the code amount is reduced.

  A packet, which is a code unit of JPEG 2000 encoded data, has attributes of component C, resolution level L, position (precinct) P, and layer (image quality) L, and packets are arranged in various orders (progression orders) according to these attributes. Can be lined up. In the case of the present invention, for example, a CRPL progression order can be used as a progression order in order to easily select a packet having a necessary resolution level of a component necessary for decoding encoded data.

  FIG. 10 schematically shows the structure of the coded data of the CRPL progression order.

《Adding information to select valid components during decryption》
In processing block 114, information for selecting a valid component according to the decoding resolution level when decoding the encoded data can be described in the header of the encoded data. For example, such information can be described as a character code in a COM marker segment in a main header added to the head of encoded data. In the case of encoded data with 5 resolution levels and 5 components as described with reference to FIGS.
”Res = 5: Comp = 0; Res = 4: Comp = 0,1; Res = 3: Comp = 0,1,2; Res = 2: Comp = 0,1,2,3:
Res = 1: Comp = 0,1,2,3,4 ”
Can be written. For example, “Res = 4: Comp = 0,1” in this description means that components C0 and C1 should be selected as effective components when decoding at decoding resolution level 4. However, such a description method is merely an example, and any other description method may be adopted.

  When such information can be given to the decoding side as information different from the encoded data, or when the correspondence between the decoding resolution level and the component to be decoded is fixed, Such information may not be described in the header of each encoded data, and such an aspect is also included in the present invention.

<< Addition of information for selecting component overlay method during decoding >>
In the processing block 114, information for selecting a component superimposing method according to the decoding resolution level when decoding the encoded data may be described in the header of the encoded data.

For example, in Part 2 of JPEG2000, component conversion and component reverse conversion by a matrix conversion method are possible. Component superimposition processing at the time of reduced decoding can be performed by inverse matrix component transformation. When the encoded data of the resolution level number 5 and the component number 5 described above and each component is a binary image, the component superimposing process for obtaining the restored image at the decoding resolution level i is as follows.
W5 = C0
W4 = C0 + C1
W3 = C0 + C1 + C2
W2 = C0 + C1 + C2 + C3
W1 = C0 + C1 + C2 + C3 + C4
("+" Means logical sum). In Part 2 of JPEG2000, such a transformation matrix for component inverse transformation can be described in the MCC marker segment of the main header. Of course, it is also possible to describe the content of this component reverse transformation as a character code in the COM marker segment of the main header.

  The above is the case where the component to be encoded is a binary image. When the component is multi-valued, that is, when the original image and the translation image are multi-valued images, the block diagram of the encoding processing unit 103 is as shown in FIG. As can be seen from a comparison between FIG. 11 and FIG. 6, a processing block 115 for performing DC level shift processing for reducing the dynamic range by half is added to each component. The rest is the same as FIG. Note that it is possible to omit the processing block 112 for setting the invalid conversion coefficient value of components other than the component C0 to 0, and such an aspect is also included in the present invention.

  Even in this case, as in the case of processing a binary image, it is natural that information for selecting an effective component according to the decoding resolution level can be described in the main header of the encoded data at the time of decoding. There are also such embodiments.

  As will be understood from the above description, the present embodiment corresponds to an embodiment of the invention according to claims 1 to 4. The processing content executed by the image processing apparatus according to this embodiment corresponds to an embodiment of the invention according to claim 11.

[Second Embodiment]
FIG. 12 is a block diagram of an image processing apparatus according to the second embodiment of the present invention. The image processing apparatus 200 includes a code input unit 201, a decoding processing unit 202, a component overlay processing unit 203, a decoding resolution level designation unit 204, and a component selection unit 205. The image processing apparatus 200 can be realized by software by using hardware resources of a computer having a general configuration including, for example, a CPU and a memory.

  The code input unit 201 is means for fetching encoded data (or similar encoded data) generated by the image processing apparatus according to the first embodiment from a storage (not shown) in which the encoded data is stored. The decoding processing unit 202 is means for performing decoding processing according to the JPEG2000 decoding algorithm on the encoded data input by the code input unit 201. The decoding resolution level designation unit 204 is a means for designating the decoding resolution level in the decoding processing unit 202. The component selection unit 205 is means for selecting an effective component according to the decoding resolution level designated by the decoding resolution level designation unit 205. The decoding processing unit 202 executes decoding processing up to the specified decoding resolution level, with only valid components selected by the component selection unit 205 as decoding targets. Therefore, the component restored by the decoding processing unit 202 is only the component selected by the component selection unit 205, and the resolution level thereof is the designated restoration resolution level.

  When the input encoded data is encoded using a binary original image and its translation image as components, the block diagram of the decoding processing unit 200 can be shown as in FIG. That is, in the processing block 211, entropy decoding by coefficient bit modeling and arithmetic decoding is performed on the packets up to the decoding resolution level of valid components of the encoded data, thereby generating transform coefficients of those components. Then, in the processing block 212, an effective component of the restoration resolution level is restored by applying a reversible discrete wavelet inverse transformation to the transformation coefficients.

  When the input encoded data is encoded using a multi-value original image and its translation image as components, the processing block diagram of the decoding processing unit 200 is as shown in FIG. The only difference from FIG. 13 is that a processing block 213 for performing DC level reverse shift processing of the restored component is added after the processing block 212.

  When the number of components of the input encoded data, the number of resolution levels, and the amount of movement of the translation image (component C1 or lower) with respect to the original image (component C0) are fixed, the component selection unit 205 is designated. An effective component can be selected only by the decoding resolution level. Such an embodiment is also included in the present invention.

  Here, encoded data generated by the image processing apparatus according to the first embodiment is input, and information for selecting an effective component in the COM marker segment in the main header according to the decoding resolution level Is described. The component selection unit 205 can refer to the information described in the COM marker segment, select a valid component at the specified decoding resolution level, and specify it to the decoding processing unit 202.

  The component superimposing processing unit 203 performs superimposing processing of the component restored by the decoding processing unit 202, so that even in the case of reduced decoding, there is no interruption of fine lines in the original image, and fine line reproducibility is good. This is means for generating a restored image. For example, as shown in FIG. 15, the component overlay processing unit 203 can be configured by a buffer storage unit 221 and a calculation unit 222. One or more components restored by the decoding processing unit 202 are temporarily stored in the buffer storage unit 221. The calculation unit 222 is a means for obtaining the pixel value of the restored image from the components stored in the buffer storage unit 221.

  In the case of reduced decoding, the buffer storage unit 221 stores two or more components. The calculation unit 222 obtains the pixel value at the corresponding position of the restored image from the pixel values at the same position of these components.

  That is, when the component is a binary image, the calculation unit 222 can use the logical sum of the pixel values at the same position as the pixel value at the corresponding position of the restored image. In a binary image, black pixels are often represented by “1” and white pixels are represented by “0”, but they may be encoded by reversing black and white. In the case of such encoded data, the calculation unit 222 can use the logical product of the pixel values at the same position of two or more components in the buffer storage unit 221 as the pixel value at the corresponding position of the restored image. As described above with reference to FIG. 9, as described above with reference to FIG. 9, it is possible to obtain a restored image with excellent reproducibility of fine lines in the original image even by reduction decoding.

  When the restored component is a multi-valued image, the calculation unit 222 can set the maximum value among the pixel values at the same position of those components as the pixel value at the corresponding position of the restored image. Although it is possible to obtain an average value of pixel values at the same position of the component and use the average value as the pixel value of the restored image, the restored image becomes a pale image as a whole.

  On the other hand, when it is not reduced decoding, that is, when the highest resolution level and the decoded resolution level of the encoded data are the same, only one component C0 is stored in the buffer storage unit 221. In this case, the calculation unit 222 outputs the pixel value of the component C0 as it is as the pixel value of the restored image. Since this restored image is a lossless image of the original image, there is no interruption of fine lines.

  As is apparent from the above description, this embodiment corresponds to an embodiment of the invention according to claims 6 and 8. The processing content executed by the image processing apparatus according to this embodiment corresponds to an embodiment of the invention according to claim 12.

[Third Embodiment]
FIG. 16 is a block diagram of an image processing apparatus according to the third embodiment of the present invention. Differences between the image processing apparatus 300 and the image processing apparatus 200 according to the second embodiment are as follows.

  The decoding processing unit 202 performs decoding processing up to the decoding resolution level specified by the decoding resolution level specifying unit 204 for all components of the input encoded data, and restores all components at the decoding resolution level. The component selection unit 205 selects an effective component corresponding to the decoding resolution level, but specifies the selected component to the component superimposition processing unit 303 instead of specifying it to the decoding processing unit 202. The component superimposition processing unit 303 performs a superimposition process on only the component specified by the component selection unit 205 in the components generated by the decoding processing unit 203 to generate a restored image. The block diagram of the component overlay processing unit 303 can be expressed as shown in FIG. 15, but a valid component is designated by the component selection unit 205 to the computation unit 222, and the computation unit 222 restores the restored image from only the designated component. The pixel value of is determined.

  The image processing apparatus 300 can also be realized by software using hardware resources of a computer having a general configuration including, for example, a CPU and a memory.

  As is apparent from the above description, this embodiment corresponds to an embodiment of the invention according to claims 7 and 8. The processing content of the image processing apparatus according to the present embodiment corresponds to an embodiment of the invention according to claim 13.

[Fourth Embodiment]
FIG. 17 is a block diagram of an image processing apparatus according to the fourth embodiment of the present invention. The image processing apparatus 400 includes a code input unit 201, a decoding processing unit 402, a decoding resolution level designation unit 204, and a component superposition method selection unit 405. The image processing apparatus 400 can also be realized by software using hardware resources of a computer having a general configuration such as a CPU and a memory.

  The decoding processing unit 402 includes component inverse conversion means using a matrix conversion method defined in Part 2 of JPEG2000, and internally executes component superposition processing by component inverse conversion. A processing block diagram of the decoding processing unit 402 is shown in FIG. The target data is encoded data obtained by encoding binary components. In processing block 411, entropy decoding of the encoded data is performed to restore component transform coefficients, and the transform coefficients restored in process block 412 are reversible discrete. Apply the inverse wavelet transform to restore the component with the resolution level. Then, a component superimposition process is performed on the component by a component inverse transform of a matrix transformation method in a processing block 413 to generate a restored image.

  The component superposition method selection unit 405 is means for selecting a component superposition method in accordance with the decoding resolution level designated by the decoding resolution level designation unit 204 and designating it to the decoding processing unit 402.

  When the number of components of encoded data to be input, the number of resolution levels, and the amount of movement of the translation image (component C1 or lower) with respect to the original image (component C0) are fixed, the component superposition method selection unit 405 specifies The component superimposing method (component inverse transform method) can be selected based only on the decoded decoding level. Such an embodiment is also included in the present invention.

  Here, as described in relation to the first embodiment, as information for selecting a component superimposing method according to the decoding resolution level, a component inverse transformation matrix is a main header of encoded data. It is assumed that it is described in the MCC marker segment or the same content is described in the COM marker segment of the main header. The component superposition method selection unit 405 refers to the information described in the MCC marker segment or the COM marker segment of the main header of the input encoded data, and performs a component inverse transform method corresponding to the restoration resolution level, that is, component superposition. The matching method is designated to the decoding processing unit 402.

  The processing block 413 of the decoding processing unit 402 generates a restored image by performing component inverse transformation using a designated component inverse transformation method. In processing block 411, entropy decoding of all components is performed. However, the processing efficiency is improved when entropy decoding is performed only on components used for component inverse transform, and such an aspect is also included in the present invention.

  As is apparent from the above description, this embodiment corresponds to an embodiment of the invention according to claims 9 and 10. The processing contents in the image processing apparatus according to the present embodiment correspond to an embodiment of the invention according to claim 14.

[Fifth Embodiment]
FIG. 19 is a block diagram for explaining a fifth embodiment of the present invention. In FIG. 19, reference numerals 500 and 600 denote image processing apparatuses according to the fifth embodiment of the present invention. JPIP is a protocol for transferring only necessary codes of JPEG2000 encoded data.
It operates as a JPIP server and a JPIP client compliant with (JPEG2000 Interactive Protocol). Hereinafter, the image processing apparatus 500 is called a JPIP server, and the image processing apparatus 600 is called a JPIP client.

  FIG. 20 shows a minimum necessary request / response sequence between the client and the server. When a code request is issued from the JPIP client 600, the JPIP server 500 that has received the code extracts the code that needs to be transmitted from the encoded data of the requested image and transmits it to the JPIP client 600. The JPIP client 600 reconstructs the received code into JPEG2000 format encoded data, and then performs decoding and other processing to display an image.

  Returning to FIG. The JPIP server 500 includes an image input unit 501, a translation image generation unit 502, an encoding processing unit 503, a storage 504, a communication unit 505, a request processing unit 506, and the like. The image input unit 501 is a means for inputting an image. The translation image generation unit 502 is a unit that generates one or a plurality of translation images obtained by translating the image (original image) input by the image input unit 501 by a predetermined amount. This corresponds to the translation image generation unit 102 (FIG. 1) according to the embodiment. The encoding processing unit 503 performs encoding processing based on the JPEG2000 encoding algorithm using the original image and the translation image as components, respectively, and generates JPEG2000 format encoded data having resolution scalability composed of a plurality of components. And corresponds to the encoding processing unit 103 (FIG. 1) according to the first embodiment. The generated encoded data is stored in the storage 504 as a file. The communication unit 505 is means for communicating with the JPIP client 600 or the like via a network such as a local area network (LAN) or the Internet. The request processing unit 506 is means for processing a request such as a code request from the client side, and details of the processing will be described later.

  The JPIP client 600 includes a communication unit 601, a request issuance processing unit 602, a code reconfiguration processing unit 603, a decoding processing unit 604, a component overlay processing unit 605, a display unit 606, and the like. The communication unit 601 is means for communicating with the JPIP server 500 or the like via a network. The request issuance processing unit 602 is a unit that performs processing for issuing a code request or the like. The code reconstruction processing unit 603 is means for performing processing for reconstructing encoded data in JPEG 2000 format from the code received by the communication unit 601. The decoding processing unit 604 is means for performing decoding processing of the reconstructed encoded data and restoring each component. The component overlay processing unit 605 is a unit that performs overlay processing of the restored components and generates a restored image. A display unit 606 is means for displaying the restored image on the display screen.

  The JPIP server 500 and the JPIP client 600 can be realized by software using hardware resources of a computer having a general configuration including, for example, a CPU, a memory, a network interface, and the like.

  The operation will be described below. In the JPIP client 602, the request issuance processing unit 602 issues a code request, and the code request is transmitted to the JPIP server 500 via the network by the communication unit 601.

This code request includes information such as the resource name (file name) of the target image and the size of the image to be displayed, called a view window request. The view window request includes the frame size (fx, fy) that gives the resolution and the size of the display area in the frame.
(sx, sy) and the display area offset (ox, oy) are given as parameters (see FIG. 21).

  FIG. 22 shows a processing flow on the JPIP server 600 side when a code request is transmitted from the JPIP client 600.

  In the JPIP server 500, when a code request is received by the communication unit 505 (step 511), the request processing unit 506 specifies encoded data to be transmitted from the resource name included in the received code request (step 512).

  The request processing unit 506 reads the size, tile size, resolution level, and number of components of the original image from the main header of the identified encoded data, and the “decoding resolution” described in the COM marker segment in the main header. Information for selecting an effective component according to the level is read (step 513).

  The request processing unit 506 determines a decoding resolution level from the original image size and the frame size (fx, fy) specified by the view window request (step 514). For example, if the original image size is “512 pixels × 256 pixels” and the frame size is “256 pixels × 128 pixels”, the decoding resolution level is one lower than the maximum resolution level.

  The request processing unit 506 determines tiles and their precincts included in the display area from the size and offset of the display area specified by the view window request (step 515).

The request processing unit 506 refers to the information read from the COM marker segment in step 513, and selects a component effective for decoding at the decoding resolution level determined in step 514 (step 516).
The request processing unit 506 extracts a code that needs to be transmitted from the encoded data (step 517). The code that needs to be transmitted is the packet of the precinct of the tile determined in step 515, and is the packet of all layers of the component selected in step 516 and having a resolution level equal to or lower than the decoding resolution level determined in step 514. (Because there is no quality parameter specification in the view window request, all layers are processed as specified).

  The request processing unit 506 transmits the extracted code to the JPIP client 600 through the communication unit 505 (step 518). For example, in the case of a precinct unit transfer method called JPIP JPP-stream, packets are transmitted in units of precincts. The main header and tile part header information including information necessary for decoding on the client side is also transmitted.

  In the JPIP client 600, the code reconfiguration processing unit 603 reconstructs encoded data in JPEG2000 format from the packet and header information received by the communication unit 601. The decoding processing unit 604 performs a decoding process on the reconstructed encoded data, and restores a component at a decoding resolution level. Since only the codes of components necessary for image restoration are received, all the components are to be decoded, and the components to be restored are only components necessary for generating a display image. The restored component overlay process is performed by the component overlay processing unit 605, and an image to be displayed in the display area is generated. In addition, when each component is a binary image, the component superimposing method can be performed by a method of calculating a logical sum of pixel values at the same position of all the components. In some cases, this can be done by a method of taking the maximum value of the pixel values at the same position of all components. The display unit 606 displays an image generated by the component overlay process in a display area on the display screen.

  JPIP has a method of transferring codes in units of tiles called JPT-Stream, and it is also possible to transmit and receive necessary codes by this method. Such a mode is naturally included in the present embodiment.

  As is clear from the above description, the JPIP server 500 according to this embodiment corresponds to an embodiment of the invention according to claim 5.

  In addition, although it demonstrated as a JPIP client server system here, it is not limited to this. For example, the decoding resolution level is designated from the client side, and only the component codes necessary for image restoration at the decoding resolution level are extracted on the server side as described with reference to FIG. It is also possible to construct a client / server system that reconfigures (but is not limited to) encoded data and transmits it to the client. Naturally, such a server is also included in the invention according to claim 5.

  The embodiment of the present invention has been described above. However, the present invention is not limited only to these embodiments. For example, the encoding algorithm is not limited to the one according to JPEG2000, and other than generating encoded data having resolution scalability consisting of a plurality of components by entropy encoding the discrete wavelet transform coefficients of each component. The encoding algorithm can also be used.

  A program for realizing the image processing apparatus according to each of the embodiments using a computer, that is, a program for causing a computer to function as each component of the image processing apparatus, and such a program are recorded. Various recording (storage) media that can be read by a computer, such as a semiconductor storage element, a magnetic disk, an optical disk, and a magneto-optical disk, are also included in the present invention.

It is a block diagram for demonstrating 1st Embodiment. It is explanatory drawing of an original image. It is explanatory drawing of the image which cut off the left end and upper end of the original image by 1 pixel. It is explanatory drawing of the image which moved the original image to the horizontal and vertical direction only 1 pixel. It is explanatory drawing of the image which moved the original image to the horizontal and vertical direction only 3 pixels. It is a block diagram of the encoding process part which processes a binary component. It is a figure which shows the subband decomposition | disassembly which applied discrete wavelet transform 5 times. It is a figure for demonstrating the relationship between the moving amount | distance of a component, and an invalid resolution level. It is a figure for demonstrating the effective resolution level for every component at the time of a decoding. It is explanatory drawing of the code structure of a CRPL progression order. It is a block diagram of the encoding process part which processes a multi-value component. It is a block diagram for demonstrating 2nd Embodiment. It is a block diagram of the decoding process part which processes the encoding data of a binary component. It is a block diagram of the decoding process part which processes the encoding data of a multi-value component. It is a block diagram of a component superimposition process part. It is a block diagram for demonstrating 3rd Embodiment. It is a block diagram for demonstrating 4th Embodiment. It is a block diagram of the decoding process part in FIG. It is a block diagram for demonstrating 5th Embodiment. It is a figure which shows the sequence of the request response between a client and a server. It is explanatory drawing of the parameter of a view window request. It is a figure for demonstrating the processing flow in a JPIP server. It is a figure which shows the binary image containing the thin line of 1 pixel width. It is a figure which shows the subband decomposition | disassembly which applied discrete wavelet transform 3 times.

Explanation of symbols

DESCRIPTION OF SYMBOLS 101 Image input part 102 Parallel moving image generation part 103 Encoding process part 201 Code input part 202 Decoding process part 203 Component superimposition process part 204 Decoding resolution level designation | designated part 205 Component selection part 303 Component superimposition process part 402 Decoding process part 405 Component superposition method selection unit 500 JPIP server 501 Image input unit 502 Parallel image generation unit 503 Encoding processing unit 504 Storage 505 Communication unit 506 Request processing unit

Claims (14)

Image input means for inputting images,
Image generating means for generating one or a plurality of translation images obtained by translating the original image from an image input by the image input means (hereinafter referred to as an original image);
An encoding process for generating encoded data having resolution scalability consisting of a plurality of components by performing discrete wavelet transform for each component, and entropy encoding the transform coefficient, with the original image and the parallel moving image as components. means,
An image processing apparatus comprising:   The encoding processing unit sets a value of a conversion coefficient equal to or higher than a resolution level determined by a translation amount of the translation image with respect to the original image, to 0 among transformation coefficients of the translation image. Item 8. The image processing apparatus according to Item 1.   The encoding processing means describes information for selecting an effective component in accordance with a decoding resolution level when decoding the encoded data in a header of the encoded data. The image processing apparatus described.   The encoding processing means describes information for selecting a component superposition method in accordance with a decoding resolution level when decoding the encoded data in a header of the encoded data. The image processing apparatus according to 1. Storage means for storing encoded data generated by the encoding processing means;
Means for communication with external devices,
Based on the code request from the external device received by the communication unit, the code that needs to be transmitted to the external device is extracted from one encoded data stored in the storage unit, and the extracted code is The image processing apparatus according to claim 1, further comprising request processing means for transmitting to an external apparatus by the communication means. An original image and one or a plurality of translated images obtained by translating the original image are used as components, respectively, and discrete wavelet transform is performed for each component, and the transform coefficient is generated by entropy encoding, and includes a plurality of components. Code input means for inputting encoded data having resolution scalability;
A decoding resolution level specifying means for specifying a decoding resolution level;
Component selecting means for selecting an effective component of the encoded data (hereinafter referred to as input encoded data) input by the encoding input means in accordance with the decoding resolution level specified by the decoding resolution level specifying means;
Decoding processing means for performing decoding processing up to the decoding resolution level specified by the decoding resolution level specifying means, with the effective component selected by the component selecting means as a decoding target for the input encoded data;
Component overlay processing means for generating a restored image by performing overlay processing of components restored by the decoding processing of the decoding processing means;
An image processing apparatus comprising: A resolution composed of a plurality of components generated by performing discrete wavelet transform for each component and entropy encoding the transform coefficient for each component, using the original image and one or a plurality of translated images obtained by translating the original image as components. Code input means for inputting encoded data having scalability;
A decoding resolution level specifying means for specifying a decoding resolution level;
Component selecting means for selecting an effective component of the encoded data (hereinafter referred to as input encoded data) input by the encoding input means in accordance with the decoding resolution level specified by the decoding resolution level specifying means;
Decoding processing means for performing decoding processing up to the decoding resolution level specified by the decoding resolution level specifying means, with all components as decoding targets for the input encoded data;
Component superimposing processing means for generating a restored image by performing superimposing processing of effective components selected by the component selecting means in the components restored by the decoding processing of the decoding processing means;
An image processing apparatus comprising: In the header of the input encoded data, information for selecting an effective component according to the decoding resolution level when the input encoded data is decoded is described.
7. The component selecting unit selects an effective component based on the information described in a header of the input encoded data and the decoding resolution level specified by the decoding resolution level specifying unit. Or the image processing apparatus according to 7. A resolution composed of a plurality of components generated by performing discrete wavelet transform for each component and entropy encoding the transform coefficient for each component, using the original image and one or a plurality of translated images obtained by translating the original image as components. Code input means for inputting encoded data having scalability;
A decoding resolution level specifying means for specifying a decoding resolution level;
Decoding processing up to the decoding resolution level designated by the decoding resolution level designation unit is performed on all the components as decoding targets for the coded data input by the code input unit (hereinafter referred to as input coded data). Decryption processing means to perform,
Component superposition method selection means for selecting a superposition method of components restored in the decoding processing of the decoding processing means according to the decoding resolution level designated by the decoding resolution level designation means;
Have
The decoding processing means includes means for generating a restored image by performing a superimposition process on the component restored by the decoding process by the superimposing method of the components selected by the component superposition method selecting means. An image processing apparatus. In the header of the input encoded data, information for selecting a component superposition method according to a decoding resolution level when the input encoded data is decoded is described.
The component superposition method selection means selects a component superposition method based on the information described in the header of the input encoded data and the decoding resolution level designated by the decoding resolution level designation means. The image processing apparatus according to claim 9. An image input process for inputting an image,
An image generation step of generating one or a plurality of parallel movement images obtained by translating the original image from the image input by the image input step (hereinafter referred to as an original image);
An encoding process for generating encoded data having resolution scalability consisting of a plurality of components by performing discrete wavelet transform for each component, and entropy encoding the transform coefficient, with the original image and the parallel moving image as components. Process,
An image processing method comprising: An original image and one or a plurality of translated images obtained by translating the original image are used as components, respectively, and discrete wavelet transform is performed for each component, and the transform coefficient is generated by entropy encoding, and includes a plurality of components. A code input process for inputting encoded data having resolution scalability;
A decoding resolution level specifying step for specifying a decoding resolution level;
A component selection step of selecting an effective component of the encoded data (hereinafter referred to as input encoded data) input by the encoding input step according to the decoding resolution level specified by the decoding resolution level specifying step;
A decoding process step of performing a decoding process up to the decoding resolution level designated by the decoding resolution level designation step, with the effective component selected by the component selection step as a decoding target for the input encoded data;
A component superimposition process step of generating a restored image by performing a superimposition process of the component restored by the decoding process of the decoding process step;
An image processing method comprising: An original image and one or a plurality of translated images obtained by translating the original image are used as components, respectively, and discrete wavelet transform is performed for each component, and the transform coefficient is generated by entropy encoding, and includes a plurality of components. A code input process for inputting encoded data having resolution scalability;
A decoding resolution level specifying step for specifying a decoding resolution level;
A component selection step of selecting an effective component of the encoded data (hereinafter referred to as input encoded data) input by the encoding input step according to the decoding resolution level specified by the decoding resolution level specifying step;
A decoding process step for performing decoding processing up to the decoding resolution level specified by the decoding resolution level specifying step, with all components as decoding targets, for the input encoded data;
A component superimposition processing step for generating a restored image by performing superimposition processing of the effective components selected by the component selection step in the components restored by the decoding processing of the decoding processing step;
An image processing method comprising: An original image and one or a plurality of translated images obtained by translating the original image are used as components, respectively, and discrete wavelet transform is performed for each component, and the transform coefficient is generated by entropy encoding, and includes a plurality of components. A code input process for inputting encoded data having resolution scalability;
A decoding resolution level specifying step for specifying a decoding resolution level;
Decoding processing up to the decoding resolution level specified in the decoding resolution level specifying step is performed on all the components as decoding targets for the encoded data input in the code input step (hereinafter referred to as input encoded data). A decryption processing step to be performed;
A component superposition method selection step of selecting a superposition method of components restored in the decoding process of the decoding processing step according to the decoding resolution level designated by the decoding resolution level designation step,
The decoding process step includes a step of generating a restored image by performing a superimposition process on the component restored by the decoding process by the superimposing method of the component selected by the component superposition method selection step. An image processing method.

Images