JP4353986B2 - Image composition apparatus, image composition method, image composition program, and computer-readable recording medium on which image composition program is recorded - Google Patents

Description

  The present invention relates to an image synthesizing apparatus, an image synthesizing method, an image synthesizing program, and a computer-readable recording medium on which an image synthesizing program is recorded, and in particular, an image synthesizing apparatus and an image synthesizing method for synthesizing a plurality of encoded images. The present invention relates to an image composition program and a computer-readable recording medium on which the image composition program is recorded.

  In recent years, opportunities for using encoded moving image data and still image data are increasing due to the digitization of television broadcasting, the distribution of moving images on the Internet, the spread of digital cameras and digital video cameras, and the like. Along with this, editing of encoded image data has been widely performed. In many cases, an irreversible encoding method is used in which orthogonal transform processing, quantization processing, and encoding processing of moving image data or still image data from the spatial domain to the frequency domain are executed in this order. In addition, in order to decode the encoded image data up to the spatial domain, a coefficient decoding process, an inverse quantization process, and an orthogonal inverse transform process from the frequency domain to the spatial domain are executed in this order. For the orthogonal transformation process, an orthogonal transformation technique such as discrete cosine transformation (DCT) or Karoonen / Labe transformation (KLT) is used. In recent years, there has been an increasing need for easy editing and processing of encoded image data of an irreversible encoding method using this orthogonal transformation technique.

  An image encoding method using orthogonal transform such as discrete cosine transform (DCT) that is widely used at present performs encoding by lossless encoding processing such as orthogonal transformation processing, quantization processing, and Huffman coding. .

  Here, image data expressed by pixel data in the spatial domain is defined as spatial domain image data, and unquantized image data obtained by orthogonal transformation processing of the spatial domain image data is defined as orthogonal transformation domain image data. Define. Also, image data obtained by quantizing the orthogonal transform area image data is defined as quantized orthogonal transform area image data, and image data obtained by encoding the quantized orthogonal transform image data is defined as orthogonal transform area encoded image data. Define.

  Here, synthesis of orthogonal transform region encoded image data will be described. FIG. 16 is a diagram for explaining a process of combining a plurality of image data. 16A shows the first orthogonal transform region encoded image data, FIG. 16B shows the second orthogonal transform region encoded image data, and FIG. 16C shows the orthogonal transform region encoded after synthesis. Image data is shown. A part of the second orthogonal transformation region encoded image data 102 (portion surrounded by a dotted line) is pasted into the first orthogonal transformation region encoded image data 101, and a new combined orthogonal transformation region encoded image data 103 is obtained. The case of creating is shown.

In the conventional synthesizing process, the synthesizing process is performed in the following procedure.
(1) The first orthogonal transform region encoded image data 101 and the second orthogonal transform region encoded image data 102 are respectively decoded, and the first quantized orthogonal transform region image data and the second quantized orthogonal transform region image data are converted into create.

  (2) The first quantized orthogonal transform region image data and the second quantized orthogonal transform region image data are inversely quantized to generate first orthogonal transform region image data and second orthogonal transform region image data.

  (3) Orthogonal inverse transform is performed on the first orthogonal transformation region image data and the second orthogonal transformation region image data to create first spatial region image data and second spatial region image data.

  (4) The portion surrounded by the dotted line 102 in FIG. 16 in the second spatial area image data is cut out and pasted to 101 in FIG. 16 in the first spatial area image data to create composite spatial area image data.

  (5) The composite space area image data is subjected to orthogonal transform processing to generate composite orthogonal transform area image data.

  (6) Quantization processing is performed using a quantization table in which the synthesized orthogonal transform area image data is designated, and synthesized quantized orthogonal transform area image data is created.

  (7) The composite quantized orthogonal transform area image data is losslessly encoded to generate composite orthogonal transform area encoded image data 103.

However, this conventional method for synthesizing image data has the following three problems.
(1) Processing speed problem In order to synthesize an image in the spatial domain, a process for transforming the orthogonal transform domain encoded image data used for the synthesis into the spatial domain image data, and a synthesized spatial domain image created in the spatial domain A process for converting from data to composite orthogonal transform area encoded image data is necessary, and the process takes a lot of time. For this reason, there exists a problem that processing speed becomes slow. In particular, when the image data is a moving image, there is a time limit when, for example, it is desired to synthesize in real time.

(2) Appropriate image quality determination takes time Since the encoding method of the composite space area image data is an irreversible encoding method, deterioration in image quality is inevitable. As for the image quality, it is necessary to visually check the synthesized spatial domain image data obtained by decoding the synthesized orthogonal transform domain encoded image data once encoded, and determine whether or not the desired image quality is obtained. As a result of confirmation, if the desired image quality is not reached, parameters such as the quantization step must be changed, and a series of processing associated with encoding after the synthesized space area image data must be performed again. In addition, it is necessary to repeat this until a desired image quality is obtained, which requires time and labor.

(3) The problem of reduction in compression rate To quantize the quantization step used to quantize the first image data, the quantization step used to quantize the second image data, and the synthesized image data The quantization steps are determined independently. For this reason, the capacity of the encoded combined image data may become large for the image quality.

  Among these problems, (2) a technique for solving the problem that it takes time to determine an appropriate image quality is described in Japanese Patent Laid-Open No. 7-184199 (Patent Document 1). Japanese Patent Laid-Open No. 7-184199 discloses encoding and decoding processing by repeatedly encoding and decoding image data a plurality of times using a fixed-length encoding method. The mean square error between the data convergence means for converging to a reversible state in which no change occurs between data and image data, and the initial image data and the pixel data after converging to the reversible state is below a predetermined level It is described that a data convergence confirmation means for confirming this is provided, and convergent data by a quasi-reversible encoding method is used as decoded data.

However, in the image data encoding circuit described in Japanese Patent Laid-Open No. 7-184199, it is possible to reduce labor for visual confirmation, but encoding processing including quantization processing and decoding processing including inverse quantization are performed. Since the process is executed a plurality of times, it takes time for the process or a high processing capacity is required for the apparatus. In addition, the image quality can be set to a desired level, but the amount of data after encoding may increase.
Japanese Patent Laid-Open No. 7-184199

  The present invention has been made to solve the above-described problems, and one of the objects of the present invention is to perform image synthesis that can prevent deterioration in image quality and compression rate when generating a composite image. Is to provide a device.

  Another object of the present invention is to provide an image synthesizing apparatus capable of reducing the load on the apparatus for synthesizing an encoded image.

  Still another object of the present invention is to provide an image composition method capable of preventing deterioration in image quality and reduction in compression rate when a composite image is generated.

  Still another object of the present invention is to provide an image composition method capable of reducing the load on the apparatus.

  Another object of the present invention is to provide an image composition program capable of preventing deterioration in image quality and reduction in compression rate when generating a composite image, and a computer-readable recording medium on which the program is recorded. is there.

  Still another object of the present invention is to provide an image composition program capable of reducing the load on the computer and a computer-readable recording medium on which the program is recorded.

  In order to achieve the above object, according to one aspect of the present invention, an image synthesizing apparatus converts a function transform and a quantized function from a spatial domain into a data represented by a function transform coefficient value by a set of basis functions. Input means for inputting one or a plurality of image data including a transform coefficient value, a quantization table corresponding to each of the one or a plurality of image data, and a combined quantization table, and the input one or a plurality of image data Extracting means for extracting a plurality of regions from the combining means, combining means for generating combined data obtained by combining the plurality of extracted regions, and function transform coefficient values of each of the plurality of extracted regions are quantized corresponding to the regions Conversion means for converting based on the table and the synthetic quantization table.

  According to the present invention, a plurality of regions extracted from one or a plurality of image data are synthesized. The function conversion coefficient values of each of the plurality of extracted regions are converted based on the quantization table and the synthesized quantization table corresponding to the regions. This conversion is performed by, for example, using a quantization table used for quantization for each of a plurality of quantized regions, and then using a single composite quantization table for the plurality of inverse quantized regions. This is the process of quantization. For this reason, the error which generate | occur | produces by quantizing the function conversion coefficient value of each area | region can be reduced. As a result, it is possible to provide an image composition device capable of preventing deterioration in image quality and reduction in compression rate when generating a composite image.

  According to another aspect of the present invention, an image synthesizing apparatus includes a function transform that converts data from a spatial domain into data represented by a function transform coefficient value by a set of basis functions, and a quantized function transform coefficient value. Or an input means for inputting a plurality of image data and a quantization table corresponding to each of the one or more image data; an extraction means for extracting a plurality of regions from the input one or more image data; A synthesis unit that generates synthesized data obtained by synthesizing the plurality of regions, a synthesis quantization table creation processing unit that creates a synthesis quantization table based on at least the quantization table corresponding to the plurality of extracted regions, Transform means for transforming the function transform coefficient values of each of the plurality of extracted regions based on a quantization table and a synthetic quantization table corresponding to the regions; Obtain.

  According to the present invention, since the synthetic quantization table is determined based on the quantization table corresponding to each of the plurality of regions, it is possible to determine the synthetic quantization table with a small error caused by quantization. For this reason, since it is not necessary to visually confirm the decoded image data, it is not necessary to repeat the decoding process and the encoding process a plurality of times. As a result, it is possible to provide an image synthesizing apparatus that can reduce the load on the apparatus that synthesizes encoded images.

  Preferably, the conversion unit converts the function conversion coefficient value of each of the plurality of extracted regions based on a corresponding element of the quantization table of the region and a corresponding element of the combined quantization table.

  According to still another aspect of the present invention, an image synthesizing apparatus includes a function transform that converts data from a spatial domain to a data represented by a function transform coefficient value by a set of basis functions, and a quantized function transform coefficient Alternatively, input means for inputting a plurality of image data, a quantization table corresponding to each of the one or more image data, and a combined quantization table, and extracting a plurality of regions from the input one or more image data Extraction means, synthesis means for generating synthesized data obtained by synthesizing the plurality of extracted areas, and function coefficient values of the extracted areas are inversely quantized using a quantization table corresponding to the areas. Inverse quantization means, and quantization means for quantizing a function transform coefficient value obtained by inverse quantization of the synthesized data using the synthesized quantization table.

  According to the present invention, a plurality of regions extracted from one or a plurality of image data are combined to generate combined data. Each of the function coefficient values of the plurality of regions is inversely quantized using the quantization table corresponding to the region, and the inversely quantized function transform coefficient value of the synthesized data is quantized using the combined quantization table. The For this reason, the error which generate | occur | produces by quantizing the function conversion coefficient value of each area | region can be reduced. As a result, it is possible to provide an image composition device capable of preventing deterioration in image quality and reduction in compression rate when generating a composite image.

  According to still another aspect of the present invention, an image synthesizing apparatus includes a function transform that converts data from a spatial domain to a data represented by a function transform coefficient value by a set of basis functions, and a quantized function transform coefficient. Alternatively, an input unit that inputs a plurality of image data and a quantization table corresponding to each of the one or more image data, an extraction unit that extracts a plurality of regions from the input one or more image data, and an extraction A synthesis means for generating synthesized data obtained by synthesizing the plurality of regions, and inverse quantization for inversely quantizing each of the function coefficient values of at least the plurality of regions extracted using a quantization table corresponding to the region Means, a synthetic quantization table creation processing means for creating a synthetic quantization table based on a quantization table corresponding to each of a plurality of regions, and the created synthetic quantization Using Buru, and a quantizing means for quantizing the dequantized function conversion coefficient value of combined data.

  According to the present invention, since the synthesized quantization table is determined based on the quantization table used when quantizing a plurality of regions, the function transform coefficient value of each region is quantized using the synthesized quantization table. By doing so, errors that occur can be reduced. In addition, since it is not necessary to repeat the decoding process and the encoding process a plurality of times, it is possible to provide an image composition apparatus that can reduce the load on the apparatus.

  Preferably, the synthetic quantization table creation processing unit determines a synthetic quantization table based on a ratio of corresponding elements of the quantization table corresponding to each of the plurality of regions.

  Preferably, the synthesized quantization table creation processing means sets each element of the synthesized quantization table as a common divisor of the corresponding elements of the quantization table corresponding to each of the plurality of regions.

  Preferably, the synthesized quantization table creation processing unit sets each element of the synthesized quantization table as the greatest common divisor of the corresponding elements of the quantization table corresponding to each of the plurality of regions.

According to the present invention, errors generated by quantization can be further reduced.
According to still another aspect of the present invention, an image composition method includes a function transform for transforming data from a spatial domain into a data represented by a function transform coefficient value by a set of basis functions, and a quantized function transform coefficient value. Inputting one or a plurality of image data, a quantization table corresponding to each of the one or a plurality of image data, and a synthetic quantization table, and extracting a plurality of regions from the input one or a plurality of image data A step, a step of generating synthesized data obtained by synthesizing the plurality of extracted regions, and a function conversion coefficient value of each of the extracted regions based on a quantization table corresponding to the region and a synthesized quantization table Converting.

  According to the present invention, it is possible to provide an image composition method capable of preventing deterioration in image quality and reduction in compression rate when generating a composite image.

  According to still another aspect of the present invention, an image synthesis method includes a function transform for transforming data from a spatial domain into data represented by a function transform coefficient value by a set of basis functions, and a quantized function transform coefficient value. A step of inputting one or a plurality of image data and a quantization table corresponding to each of the one or a plurality of image data; a step of extracting a plurality of regions from the input one or a plurality of image data; Generating synthesized data obtained by synthesizing the plurality of areas, creating a synthesized quantization table based on at least the quantization table corresponding to the extracted areas, and each of the extracted areas Transforming the function transform coefficient value based on a quantization table corresponding to the region and a synthesized quantization table.

  According to the present invention, it is possible to provide an image composition method capable of preventing deterioration in image quality and reduction in compression rate. Further, it is possible to provide an image composition method that can reduce the load on the apparatus.

  According to still another aspect of the present invention, an image composition method includes a function transform for transforming data from a spatial domain into data represented by a function transform coefficient value by a set of basis functions, and a quantized function transform coefficient value. Inputting one or a plurality of image data and a quantization table corresponding to each of the one or a plurality of image data and a synthetic quantization table, and extracting a plurality of regions from the input one or a plurality of image data A step of generating combined data obtained by combining the plurality of extracted regions, and a step of inversely quantizing each of the function coefficient values of the extracted regions using a quantization table corresponding to the regions; And quantizing the function transform coefficient value obtained by dequantizing the synthesized data using the synthesized quantization table.

  According to the present invention, it is possible to provide an image composition method capable of preventing deterioration in image quality and reduction in compression rate when generating a composite image.

  According to still another aspect of the present invention, an image composition method includes a function transform for transforming data from a spatial domain into a data represented by a function transform coefficient value by a set of basis functions and a quantized function transform coefficient. Or inputting a plurality of image data and a quantization table corresponding to each of the one or more image data, extracting a plurality of regions from the input one or more image data, and Generating composite data by combining a plurality of areas, dequantizing at least the extracted function coefficient values of the plurality of areas using a quantization table corresponding to the areas, and each of the plurality of areas A step of creating a synthetic quantization table based on a quantization table corresponding to, and using the created synthetic quantization table, And a step of quantizing the function conversion coefficients quantized.

  According to the present invention, it is possible to provide an image composition method capable of preventing deterioration in image quality and reduction in compression rate. Further, it is possible to provide an image composition method that can reduce the load on the apparatus.

  According to still another aspect of the present invention, an image composition program causes a computer to execute each step of the image composition method.

  According to the present invention, it is possible to provide an image composition program capable of preventing image quality deterioration and compression rate reduction. Further, it is possible to provide an image composition program that can reduce the load on the computer.

  According to still another aspect of the present invention, there is provided a computer-readable recording medium on which the image composition program is recorded.

  According to the present invention, it is possible to provide a computer-readable recording medium on which an image composition program capable of preventing image quality deterioration and compression rate reduction is recorded. Further, it is possible to provide a computer-readable recording medium that records an image composition program that can reduce the load on the computer.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings. In the following description, the same parts are denoted by the same reference numerals. Their names and functions are also the same. Therefore, detailed description thereof will not be repeated.

  The image synthesizing apparatus according to the present embodiment synthesizes encoded image data encoded by an image encoding method using function conversion. Here, function conversion refers to conversion from the spatial domain using a set of basis functions. An area that is function-converted with respect to the space area is referred to as a function conversion area. Function transformation includes DCT (Discrete Cosine Transform), DFT (Discrete Fourier Transform), Walsh Hadamard Transform (WHT), Discrete Sine Transform (DST), Haar Transform, Slant Transform, Karhunen / Labe Transform (KLT), Orthogonal Wavelet It is not limited to orthogonal transformation represented by transformation, and non-orthogonal transformation such as non-orthogonal wavelet is also included.

  Here, other terms used in this specification are defined. First, image data expressed in the spatial domain is referred to as spatial domain image data. Image data obtained by function conversion is called function conversion area image data, and its coefficient is called function conversion coefficient. The value of the function conversion coefficient is referred to as a function conversion coefficient value.

  Also, image data in which the function conversion coefficient of the function conversion area image data is quantized is referred to as quantization function conversion area image data, and the coefficient is referred to as a quantization function conversion coefficient. Also, the quantization function transform area image data that is losslessly encoded such as Huffman encoding is referred to as encoded image data.

  The image data includes data representing all images, such as spatial domain image data, function transformation domain image data, quantization function transformation domain image data, and encoded image data.

In addition, conversion from a function conversion area to a space area by a set of basis functions is called function inversion.
Here, the image composition apparatus of the present embodiment will be described with reference to FIG. In the image composition apparatus according to the present embodiment, the function conversion processing for converting from the space area to the function conversion area is performed in units of blocks (8 pixels × 8 pixels), for example. Also, various coefficients and pixel values of image data are cut and pasted in units of the blocks. The size of the block may be arbitrary. When function conversion is performed in units of blocks of N pixels × M pixels, various coefficients and pixel values may be cut and pasted in units of N × M blocks.

<First Embodiment>
FIG. 1 is a functional block diagram illustrating an outline of functions of the image composition apparatus according to the first embodiment. The image composition device 1004 can be realized by a general computer. Since the hardware configuration of a computer is well known, description thereof will not be repeated here.

  Referring to FIG. 1, an image composition device 1004 is connected to a first image input device 1001, a second image input device 1002, a composite map input device 1003, and an encoded composite image data output device 1005. The image composition apparatus 1004 receives the first encoded image data from the first image input apparatus 1001, the second encoded image data from the second image input apparatus 1002, and the composition map data from the composition map input apparatus 1003. Is entered. The first encoded image data is data including a first quantization table expressing a quantization width in header information. The second encoded image data is data including the second quantization table expressing the quantization width in the header information.

  The image synthesis device 1004 generates encoded composite image data based on the first encoded image data, the second encoded image data, and the composite map data, and outputs the encoded composite image data to the encoded composite image data output device 1005.

  A CD-ROM (Compact Disc Read Only Memory) 1021 can be attached to the image composition device 1004. The image composition device 1004 may execute an image composition program recorded on a recording medium such as a CD-ROM by a CPU (Central Processing Unit). Generally, such an image composition program is read from a recording medium by a CD-ROM drive or the like and temporarily stored in a hard disk. Further, it is read from the hard disk to a random access memory (RAM) and executed by the CPU.

  Recording media are not limited to CD-ROMs and hard disks, but are flexible disks, cassette tapes, optical disks (MO (Magnetic Optical Disc) / MD (Mini Disc) / DVD (Digital Versatile Disc)), IC cards (memory cards). Medium) that carries a fixed program, such as a semiconductor memory such as an optical card, mask ROM, EPROM, EEPROM, flash ROM, or the like.

  The program here includes not only a program directly executable by the CPU but also a program in a source program format, a compressed program, an encrypted program, and the like.

  Here, the composite map data will be described. The composite map data is data that defines a method for generating a composite image. The composite image is determined by arranging at least a partial region of one or a plurality of images. Therefore, the composite map data defines the position of the region to be combined in the image that is the source of the combination and the position of the region in the combined image. Specifically, the position of the region used for the composite image in the first image data and the position of the region in the composite image, the position of the region used for the composite image in the second image data, and the composite image of the region Define the position of.

  FIG. 2 is a diagram illustrating an example of the composite map data. The composite map data will be described with reference to FIG. The composite map data includes a header part and a data part. The data portion includes first image cut area data, second image cut area data, and composite image pasting area data. The header part is data for defining an offset value of each area data. Each area data can be accessed based on the offset value defined in the header part.

  Here, composite map data corresponding to the composite processing of a plurality of image data shown in FIG. 16 will be described. FIG. 3 is a diagram showing the first image cut-out area data. 4 is a diagram showing the second image cut-out region data, and FIG. 5 is a diagram showing the composite image pasting region data. The first image cut region data corresponds to the first orthogonal transform region encoded image data 101 and has the same size as the first orthogonal transform region encoded image data 101. The second image cut-out area data corresponds to the second orthogonal transform area encoded image data 102 and has the same size as the second orthogonal transform area encoded image data 102. The combined image pasting area data corresponds to the combined orthogonal transform area encoded image data 103 and has the same size as the combined orthogonal transform area encoded image data 103. The hatched portion 201 of the first image cut-out region data and the shaded portion 401 of the composite image pasting region data all indicate the same value (for example, 1). The hatched portion 301 of the second image cut-out region data and the shaded portion 402 of the composite image pasting region data indicate that the hatched portion 201 and the hatched portion 401 are different values and all have the same value (for example, 2). The blank area 202 of the first image cut-out area data and the blank area 302 of the second image cut-out area data all indicate the same value (for example, 0).

  The area 401 in the composite image pasting area data corresponds to the hatched portion 201 in the first image cut-out area data, the area 402 in the composite image pasting area data corresponds to the hatched section 301 in the second image cut-out area data, Each region used in the synthesis is shown. A blank area 202 in the first image cut-out area data and a blank area 302 in the second image cut-out area data indicate areas that are not used for composition.

  The first image cut-out area data defines the position of an area (shaded portion 201) used for synthesis in the first encoded image data. The second image cut-out area data defines the position of an area (shaded portion 301) used for composition in the second encoded image data. The composite image pasting area data indicates the position in the composite image of the area (shaded portion 201) used for composition in the first encoded image data and the region (shaded section 301) used for composition in the second encoded image data. Define. Specifically, the region of the first encoded image data corresponding to the shaded portion 201 in the first image cut-out region data is a composite image in which the position in the encoded composite image data has the same value as the shaded portion 201. It is defined by an area 401 in the pasting area data. The region of the second encoded image data corresponding to the hatched portion 301 in the second image cut-out region data is the same as the region in the combined image pasting region data whose position in the encoded combined image data has the same value as the hatched portion 301. It is defined in area 402.

  FIG. 6 is a diagram for explaining another example of the image composition process. 6A shows the first encoded image data, FIG. 6B shows the second encoded image data, and FIG. 6C shows the encoded combined image data. FIG. 7 is another diagram showing the first image cut-out area data. FIG. 8 is another diagram showing the second image cut-out area data. FIG. 9 is another diagram showing the composite image pasting area data.

  The first image cut-out area data corresponds to the first encoded image data shown in FIG. 6A and has the same size as the first encoded image data. The second image cut-out area data corresponds to the second encoded image data shown in FIG. 6B and has the same size as the second encoded image data. The composite image pasting area data corresponds to the encoded composite image data shown in FIG. 6C and has the same size as the encoded composite image data. The hatched portion 601 of the first image cut-out region data and the shaded portion 801 of the composite image pasting region data all indicate the same value (for example, 1). The hatched portion 701 of the second image cut-out region data and the hatched portion 802 of the composite image pasting region data indicate that the hatched portion 602 and the hatched portion 801 are different values and all have the same value (for example, 2). The blank area 601 of the first image cut-out area data and the blank area 702 of the second image cut-out area data all indicate the same value (for example, 0).

  The area 801 in the composite image pasting area data corresponds to the hatched part 602 in the first image cut-out area data, the area 802 in the composite image pasting area data corresponds to the hatched part 701 in the second image cut-out area data, Each region used in the synthesis is shown. A blank area 601 in the first image cut-out area data and a blank area 702 in the second image cut-out area data indicate areas that are not used for composition.

  The first image cut-out area data defines the position of an area (shaded portion 602) used for synthesis in the first encoded image data. The second image cut-out area data defines the position of an area (shaded portion 701) used for composition in the second encoded image data. The composite image pasting region data is included in the composite encoded image data of the region (shaded portion 602) used for composition in the first encoded image data and the region (shaded portion 701) used for composition in the second encoded image data. Define the position of. Specifically, the region of the first encoded image data corresponding to the hatched portion 602 in the first image cut-out region data is a composite image in which the position in the encoded combined image data has the same value as the hatched portion 602. It is defined by an area 801 in the pasting area data. The region of the second image encoded data corresponding to the hatched portion 701 in the second image cut-out region data is the same as the region in the combined image pasting region data whose position in the encoded combined image data has the same value as the hatched portion 701. It is defined in area 802.

  The composite map data may be in any format as long as the area to be clipped and the position where the clipped area is pasted in the composite encoded image data are uniquely defined.

  Returning to FIG. 1, the image composition device 1004 receives the first encoded image data buffer 1006 for storing the first encoded image data input from the first image data input device 1001 and the second image data input device 1002. A second encoded image data buffer 1007 for storing the input second encoded image data; a composite map data buffer 1008 for storing the composite map data input from the composite map input device 1003; and first and second codes Quantization function transform coefficient that decodes quantized image data and takes out each quantized function transform coefficient of the decoded quantized function transform area image data as first quantized function transform coefficient data and second quantized function transform coefficient data An acquisition processing unit 1009 and first and second storage units for storing the extracted first and second quantization function transform coefficient data, respectively; 2 quantization function transform coefficient data buffers 1011 and 1013, a quantization table acquisition processing unit 1010 that extracts the first and second quantization tables from the first and second encoded image data, and stores the extracted quantization tables First and second quantization table buffers 1012 and 1015, a synthesized quantization table creation processing unit 1017 that creates a synthesized quantization table from the first and second quantization tables, and first and second quantizations A synthesized quantization function transformation coefficient data creation processing unit 1014 for creating the first and second quantization function transformation coefficient data by converting the function transformation coefficient data, respectively, and storing the created synthesized quantization function transformation coefficient data A synthesized quantization function transform data buffer 1016, a synthesized quantization function transform coefficient data, and a synthesized quantization data Includes a coded composite image data creation processing unit 1019 for creating a coded composite image data from the table, the encoded composite image data buffer 1020 for storing the encoded composite image data.

  First encoded image data buffer 1006, second encoded image data buffer 1007, composite map data buffer 1008, first and second quantization function transform coefficient data buffers 1011 and 1013, and first and second quantization table buffer 1012 , 1015, the combined quantization table buffer 1018, the combined quantization function transform coefficient data buffer 1016, and the encoded combined image data buffer 1020 are realized by a random access memory (RAM) such as a flash memory or a hard disk.

  A quantization function conversion coefficient acquisition processing unit 1009, a quantization table acquisition processing unit 1010, a synthetic quantization table creation processing unit 1017, a synthetic quantization function transformation coefficient data creation processing unit 1014, and an encoded synthesized image data creation processing unit 1019 For example, each is realized by an independent circuit. Further, for example, a virtual circuit realized by an arithmetic processing circuit such as a computer may be used.

  The quantization function transform coefficient acquisition processing unit 1009 reads and decodes the first encoded image data from the first encoded image data buffer 1006, and the first quantization function transform coefficient data is acquired. Then, the first quantization function conversion coefficient data is stored in the first quantization function conversion coefficient data buffer 1011. The quantization function transform coefficient acquisition processing unit 1009 reads the second encoded image data from the second encoded image data buffer 1007 and decodes it. As a result, the second encoded image data is decoded and second quantization function transform coefficient data is acquired. Then, the second quantization function conversion coefficient data is stored in the second quantization function conversion coefficient data buffer 1013.

  The quantization table acquisition processing unit 1010 reads the first encoded image data from the first encoded image data buffer 1006 and extracts the first quantization table. Then, the extracted first quantization table is stored in the first quantization table buffer 1012. Also, the quantization table acquisition processing unit 1010 reads the second encoded image data from the second encoded image data buffer 1007 and extracts the second quantization table. Then, the extracted second quantization table is stored in the second quantization table buffer 1015.

  The synthetic quantization table creation processing unit 1017 reads the first quantization table from the first quantization table buffer 1012 and the second quantization table from the second quantization table buffer 1015 to create a synthetic quantization table. Then, the generated synthesized quantization table is stored in the synthesized quantization table buffer 1018. The synthesized quantization table created by the synthesized quantization table creation processing unit 1017 is preferably a common divisor of the elements of the first quantization table and the second quantization table, each element corresponding to the element. More preferably, it is the greatest common divisor.

  The synthetic quantization function conversion coefficient data creation processing unit 1014 receives the first quantization function conversion coefficient data from the first quantization function conversion coefficient data buffer 1011 and the second quantization function from the second quantization function conversion coefficient data buffer 1013. The transformation coefficient data is read from the first quantization table buffer, the first quantization table, the second quantization table buffer from the second quantization table, and the synthesized quantization table buffer 1018 from the synthesized quantization table. Then, the synthetic quantization function transform coefficient data creation processing unit 1014 cuts out the first region specified by the first image cut-out region data of the composite map data from the first quantization function transform coefficient data, and the second image cut-out region data. The second region specified by (2) is cut out from the second quantization function transform coefficient data. Then, each of the first area and the second area is pasted in a block unit to the area specified by the composite image pasting area data of the composite map data. At the time of this pasting, the first quantization function conversion coefficient data is converted based on the corresponding first quantization data and synthesized quantization data, and the second quantization function conversion coefficient data is corresponding to it. Conversion is performed based on the second quantized data and the synthesized quantized data.

  The encoded combined image data creation processing unit 1019 reads the combined quantization function conversion coefficient data from the combined quantization function conversion coefficient data buffer 1016 and the combined quantization table from the combined quantization table buffer 1018. The encoded combined image data creation processing unit 1019 performs lossless encoding processing such as Huffman encoding on the read combined quantization function transform coefficient data. Then, encoded combined image data including data obtained by performing lossless encoding processing on the combined quantization function transform coefficient data including the combined quantization table and the image size information in the header information is created. Here, an example is shown in which the encoded combined image data has header information including a combined quantization table and image size information. However, the present invention is not limited to this, and may be data conforming to the format as encoded image data. That's fine. The encoded combined image data creation processing unit 1019 stores the generated encoded combined image data in the encoded combined image data buffer 1020.

  Here, a specific example of the synthetic quantization table will be described. FIG. 10 is a diagram illustrating an example of the first quantization table. FIG. 11 is a diagram illustrating an example of the second quantization table. FIG. 12 is a diagram illustrating an example of the synthetic quantization table. In the synthesized quantization table, each element is the greatest common divisor of the elements of the first quantization table and the second quantization table corresponding to the element.

  Next, the conversion process executed by the synthetic quantization function conversion coefficient data creation processing unit 1014 will be described. The synthesized quantization function transform coefficient data creation processing unit 1014 converts the elements of each block of the first quantization function transform coefficient data into the elements of each block of the second quantization function transform coefficient data using the following equation (1). The element value is converted using the following equation (2).

Element value of each block of first quantization function transform coefficient data = (original element value of each block of first quantization function transform coefficient data) × (corresponding element value of first quantization table data) / (synthesis Corresponding element value of quantization table) (1)
Element value of each block of second quantization function transform coefficient data = (original element value of each block of second quantization function transform coefficient data) × (corresponding element value of second quantization table data) / (synthesis Corresponding element value of quantization table) (2)
As described above, the synthetic quantization table created by the synthetic quantization table creation processing unit 1017 is a common divisor of elements of the first quantization table and the second quantization table, each element corresponding to the element, more preferably It was the greatest common divisor. By setting the common divisor, the above formulas (1) and (2) are divisible. That is, no calculation error (quantization error) occurs in the synthetic quantization function transform coefficient data creation processing unit 1014. For this reason, the encoded combined image data is not deteriorated at all. Further, by setting the greatest common divisor, the quantization function transform coefficient of the synthesized quantization function transform region image data can be minimized under the condition that the coded synthesized image data is not deteriorated at all. As a result, the quantization function transform coefficient is shifted to a low value. In general, lossless encoding of quantization function transform coefficient values is often set so that the compression rate increases when the appearance rate of low values is high, and a low coefficient value leads to an improvement in compression rate.

  FIG. 13 is a flowchart illustrating a flow of processing executed by the image composition device 1004 according to the first embodiment. Referring to FIG. 13, the image composition device 1004 stores the first encoded image data input from the first image input device 1001 in the first encoded image data buffer 1006 (S1102). Then, the second encoded image data input from the second image input device 1002 is stored in the second encoded image data buffer 1007 (S1103). Further, the composite map data input from the composite map input device 1003 is stored in the composite map data buffer 1008 (S1104).

  Next, the quantization function transform coefficient acquisition processing unit 1009 reads and decodes the first encoded image data from the first encoded image data buffer 1006 (S1105). The first quantized function transform coefficient data is obtained by decoding the first encoded image data. The extracted first quantization function transform coefficient data is stored in the first quantization function transform coefficient data buffer 1011 (S1106). The quantization table acquisition processing unit 1010 reads the first encoded image data from the first encoded image data buffer 1006 and extracts the first quantization table (S1107). The extracted first quantization table is stored in the first quantization table buffer 1012 (S1108).

  The quantization function transform coefficient acquisition processing unit 1009 further reads and decodes the second encoded image data from the second encoded image data buffer 1007 (S1109). By decoding the second encoded image data, second quantization function transform coefficient data is obtained. The extracted second quantization function transform coefficient data is stored in the second quantization function transform coefficient data buffer 1013 (S1110). The quantization table acquisition processing unit 1010 reads the second encoded image data from the second encoded image data buffer 1007 and extracts the second quantization table (S1111). Then, the extracted second quantization table is stored in the second quantization table buffer 1015 (S1112).

  Next, the synthesized quantization table creation processing unit 1017 reads the first quantization table from the first quantization table buffer 1012 and the second quantization table from the second quantization table buffer 1015, and uses them as a synthesized quantization table. Is created (S1113). Here, each element of the synthesized quantization table is the greatest common divisor of the corresponding elements of the first quantization table and the second quantization table. The created synthetic quantization table is stored in the synthetic quantization table buffer 1018 (S1114).

  Then, the synthesized quantization function transform coefficient data creation processing unit 1014 creates synthesized quantization function transform coefficient data (S1115). The synthetic quantization function conversion coefficient data creation processing unit 1014 cuts out the first area specified by the first image cut area data of the composite map data from the first quantization function conversion coefficient data and specifies the second image cut area data. The second region to be processed is cut out from the second quantization function transform coefficient data. Then, each of the first area and the second area is pasted in block units to the area specified by the composite image pasting area data of the composite map data. At the time of this pasting, the first quantization function conversion coefficient data is converted according to the above equation (1), and the second quantization function conversion coefficient data is converted according to the above equation (2). As a result, the first quantized function transform coefficient data in the first area and the quantized function transform coefficient converted data, and the second quantized function transform coefficient data in the second area and transformed data. Synthetic quantization function transform coefficient data consisting of

  The generated synthesized quantization function transform coefficient data is stored in the synthesized quantization function transform coefficient data buffer 1016 (S1116). The encoded synthesized image data creation processing unit 1019 reads the synthesized quantization function coefficient data from the synthesized quantization function transform coefficient data buffer 1016, and performs lossless coding processing such as Huffman coding on the synthesized quantization function transform coefficient data. . Then, encoded combined image data including a combined quantization table, header information such as image size information, and data obtained by lossless encoding processing of the combined quantization function transform coefficient data is created (S1117). The encoded composite image data may be in any format as long as the data conforms to the format of the encoded encoded image data. The generated encoded combined image data is stored in the encoded combined image data buffer 1020 (S1118).

<When combining three or more encoded images>
The image synthesizing apparatus 1004 described above has described an example in which two pieces of encoded image data are combined for the sake of simplicity. However, when a plurality of pieces of encoded image data are combined, they can be combined similarly. . When N pieces of encoded image data (Image1, Image2, Image3, Image4,..., ImageN) are combined, the combined quantization table data is obtained by obtaining a quantization table for each of the N pieces of encoded image data. The common divisor (preferably the greatest common divisor) of each element of the quantization table may be the element.

  Further, the conversion of the quantization function conversion coefficient data (n) included in the encoded image data (n) is performed by the following equation (3) using the quantization table (n) corresponding to the encoded image data (n). Is used.

(Element value of each block of quantization function transform coefficient of Image (n)) = (Original element value of each block of quantization function transform coefficient data of Image (n)) × (Quantization table of Image (n)) Corresponding element value of data (n) / (corresponding element value of synthetic quantization table (n)) (3)
<If the image to be combined contains an uncoded image>
When unencoded image data is included in the image to be combined, the combined quantization table creation processing unit 1017 creates a combined quantization table from the coded image quantization table. What is necessary is just to produce the quantization function conversion coefficient of the image which is not encoded using a synthetic | combination quantization table after function conversion. In the case of synthesizing one piece of encoded image data and one or more pieces of unencoded image data, the synthesized quantization table may be the same as the quantization table of encoded image data.

  As described above, the image synthesis apparatus 1004 according to the first embodiment copies a coded function transform coefficient of a plurality of coded image data in units of regions subjected to function transformation, thereby encoding a synthesized image. Create data. When the quantization function transform coefficient is copied in units of regions subjected to function transformation, a quantization table corresponding to the copy source encoded image data and a synthesis quantization table corresponding to the copy destination encoded composite image data; Convert based on Since the composition processing is performed in the function transformation region, processing such as inverse quantization, inverse function transformation, synthesized image data quantization, and function transformation processing is not required, and the number of processes can be reduced to reduce the load on the image composition device 1004. . Further, the compression rate does not decrease due to repeated quantization with a plurality of different quantization tables.

  Furthermore, each element of the quantization table corresponding to the encoded combined image data after combining is a common divisor of the corresponding elements of the quantization table corresponding to a plurality of encoded image data used for combining (in particular, the greatest common divisor is desirable). ). For this reason, since the encoded combined image data after combining does not cause any deterioration in image quality as compared with the encoded image data before combining, it is possible to determine a combined quantization table with less errors caused by quantization. it can. In particular, when the greatest common divisor is used, the quantized value of the function transform coefficient is minimized under the condition that there is no data loss due to quantization. For this reason, the number of bits required to store the quantized coefficients can be reduced, and the compression rate can be increased. Further, the range is narrowed to a range where all the values of the coefficients are all low, and this is advantageous in entropy particularly when a data string across the coefficients is losslessly encoded.

  Furthermore, since it is not necessary to visually check the decoded image data, there is no need to repeat the decoding process and the encoding process a plurality of times.

  Further, the portion 1022 in FIG. 1 is the central portion of the image coding apparatus 1004 of the present embodiment, and 1022 may be the image coding apparatus. At this time, the input of the image coding apparatus is the first quantization function transform coefficient data, the second quantization function transform coefficient data, the corresponding quantization table, and the composite map data. It is obvious that the quantization function conversion coefficient data and the corresponding quantization table may be given as quantization function conversion coefficient data.

<Second Embodiment>
FIG. 14 is a functional block diagram illustrating an outline of functions of the image composition device 1504 according to the second embodiment. The image composition device 1504 in the second embodiment includes first encoded image data input from the first image data input device 1001, second encoded image data input from the second image data input device 1002, and Based on the composite map data input from the composite map input device 1003, encoded composite image data is generated and output to the encoded composite image data output device 1005.

  The image composition device 1504 stores a first data buffer 1006 that stores first encoded image data input from the first image input device 1001, and second encoded image data input from the second image input device 1002. The second encoded image data buffer 1007, the composite map data buffer 1008 for storing the composite map data input from the composite map input device 1003, and the first and second encoded image data by decoding the first and second encoded image data. An image decoding processing unit 1509 for extracting the two spatial area image data; first and second spatial area image data buffers 1511 and 1513 for storing the extracted first and second spatial area image data; A quantization table acquisition processing unit 1010 that extracts the first and second quantization tables from the two encoded image data; First and second quantization table buffers 1012 and 1015 for storing the generated quantization table, a synthesized quantization table creation processing unit 1017 for creating a synthesized quantization table from the first and second quantization tables, A spatial region composite image data creation processing unit 1514 for creating spatial region composite image data obtained by combining the first and second spatial region image data, and a spatial region composite image for storing the created spatial region composite image data A data buffer 1516, an image encoding processing unit 1519 for generating encoded composite image data from the spatial domain composite image data and the composite quantization table, and an encoded composite image data buffer 1020 for storing the encoded composite image data Including.

  The first space area image data buffer 1511, the second space area image data buffer 1513, and the space area composite image data buffer 1516 are realized by a RAM (Random Access Memory) such as a flash memory or a hard disk.

  The image decoding processing unit 1509 and the space amount region combined image data creation processing unit 1514 are realized by, for example, independent circuits. Further, for example, a virtual circuit realized by an arithmetic processing circuit such as a computer may be used.

  The image decoding processing unit 1509 reads the first encoded image data from the first encoded image data buffer 1006, performs decoding processing, inverse quantization processing, and function inverse transformation processing on the first encoded image data, and performs first processing. Spatial area image data is generated. In the inverse quantization process, the first quantization table included in the first encoded image data is used. The generated first spatial region image data is stored in the first spatial region image data buffer 1511. In addition, the image decoding processing unit 1509 reads the second encoded image data from the second encoded image data buffer 1007, performs a decoding process, an inverse quantization process, and a function inverse transform process on the second encoded image data. Second spatial region image data is generated. In the inverse quantization process, the second quantization table included in the second encoded image data is used. The generated second space area image data is stored in the second space area image data buffer 1513.

  The spatial area composite image data creation processing unit 1514 receives composite map data from the composite map data buffer 1008, first spatial area image data from the first spatial area image data buffer 1511, and second from the second spatial area image data buffer 1513. Read the spatial domain image data. Then, the spatial domain composite image data creation processing unit 1514 creates spatial domain composite image data according to the composite map data. The spatial region composite image data creation processing unit 1514 cuts the first region specified by the first image cut region data of the composite map data from the first spatial region image data, and the second region specified by the second image cut region data. A region is cut out from the second spatial region image data. Then, each of the first area and the second area is pasted in block units to the area specified by the composite image pasting area data of the composite map data. As a result, spatial domain composite image data is created. The spatial region composite image data creation processing unit 1514 stores the created spatial region composite image data in the spatial region composite image data buffer 1516. In other words, the spatial region composite image data creation processing unit 1514 creates spatial region composite image data in the spatial region from the first spatial region image data and the second spatial region image data.

  The image encoding processing unit 1519 reads the spatial domain synthesized image data from the spatial domain synthesized image data buffer 1516 and the synthesized quantization table from the synthesized quantization table buffer 1018, respectively. Then, lossless encoding processing such as function conversion processing, quantization processing, and Huffman encoding is performed on the spatial domain composite image data. The encoded image processing unit 1519 includes a header information including a composite quantization table and image size information, and a code including data obtained by performing function conversion processing, quantization processing, and encoding processing on the spatial domain composite image data. Creates synthesized image data. Here, an example is shown in which the encoded combined image data uses the combined quantization table, image size information, and the like as header information. The image encoding processing unit 1519 stores the generated encoded combined image data in the encoded combined image data buffer 1020.

  FIG. 15 is a flowchart illustrating a flow of image composition processing executed by the image composition apparatus according to the second embodiment. Referring to FIG. 15, steps S1602 to S1604 are the same as steps S1102 to S1104 in FIG. 13, steps S1605 to S1608 are the same as steps S1107 to S1110 in FIG. 13, and step S1610 is the same as step S1114 in FIG. Therefore, description will not be repeated here.

  The image decoding processing unit 1509 reads the first encoded image data from the first data buffer 1006, decodes the first encoded image data, and generates first spatial region image data (S1611). Then, the generated first spatial area image data is stored in the first spatial area image data buffer 1511 (S1612).

  Next, the image decoding processing unit 1509 reads the second encoded image data from the second data buffer 1007, decodes the second encoded image data, and generates second spatial region image data (S1613). Then, the generated second space area image data is stored in the second space area image data buffer 1513 (S1614).

  Then, the spatial region composite image data creation processing unit 1514 receives the composite map data from the composite map data buffer 1508, the first spatial region image data from the first spatial region image data buffer 1511, and the second spatial region image data buffer 1513. The second spatial region image data is read, and spatial region composite image data is created according to the composite map data (S1615). Then, the generated composite image data is stored in the spatial domain composite image data buffer 1516 (S1616).

  The image encoding processing unit 1519 reads the spatial domain composite image data from the spatial domain composite image data buffer 1516 and the composite quantization table from the composite quantization table buffer 1018, and creates encoded composite image data (S1617). More specifically, a function conversion is performed on the spatial domain composite image data to generate a composite function conversion coefficient, and a quantization function conversion coefficient is generated by quantization of the composite function conversion coefficient using the composite quantization table. Encode function transform coefficients. The generated encoded composite image data is stored in the encoded composite image data buffer 1020 (S1618).

<When combining three or more encoded images>
The image synthesizing apparatus 1504 in the second embodiment described above has been described with respect to an example in which two encoded image data are combined to simplify the description, but there may be a case where three or more encoded image data are combined. Can be synthesized similarly. When N pieces of encoded image data (Image1, Image2, Image3, Image4,..., ImageN) are combined, the combined quantization table data obtains a quantization table for each of the N encoded images, and N pieces of encoded table data are obtained. The common divisor (preferably the greatest common divisor) of each element of the quantization table may be the element.

<If the image to be combined contains an uncoded image>
When unencoded image data is included in the image to be combined, the combined quantization table creation processing unit 1017 creates a combined quantization table from the coded image quantization table. What is necessary is just to produce the quantization function conversion coefficient of the image which is not encoded using a synthetic | combination quantization table after function conversion. Further, in the case of combining one piece of encoded image data and one or more pieces of unencoded image data, the quantization table of encoded image data may be used as it is as the combined quantization table.

  As described above, the image synthesizing device 1504 in the second embodiment decodes the first and second encoded image data, synthesizes the images in the spatial domain, and the spatial domain synthesized image data obtained by synthesizing. Are subjected to function conversion processing, quantization processing, and encoding processing. The combined quantization table used in the quantization process is a common divisor (particularly, the greatest common divisor is desirable) of each corresponding element of the quantization table corresponding to each of the first and second encoded image data. For this reason, compared with the first and second image data, the synthesized image data is not accompanied by image quality deterioration in principle. Actually, there is a possibility that slight image quality degradation may occur due to calculation errors in encoding and decoding processes, but the problem of image quality degradation can be alleviated.

  Furthermore, the synthesized quantization table used in the quantization process is a common divisor (particularly, the greatest common divisor is desirable) of each corresponding element of the quantization table corresponding to each of the first and second encoded image data. In principle, data loss due to quantization is zero. In particular, when the greatest common divisor is used, the quantized value of the function transform coefficient is minimized under the condition that there is no data loss due to quantization. For this reason, the number of bits required to store the quantized coefficients can be reduced, and the compression rate can be increased. Further, the range is narrowed to a range where all the values of the coefficients are all low, and this is advantageous in entropy particularly when a data string across the coefficients is losslessly encoded.

  Furthermore, since it is not necessary to visually check the decoded image data, there is no need to repeat the decoding process and the encoding process a plurality of times. For this reason, the load of the image composition device can be reduced.

  Note that the image composition apparatuses 1004 and 1504 in the first or second embodiment are applicable as long as they are image coding systems using function conversion. In addition, the moving image coding method is the same as the still image coding technique in terms of function conversion, quantization, and coding. For this reason, the image synthesizing apparatuses 1004 and 1504 in the first and second embodiments can be directly applied to the synthesis of moving images. Therefore, for example, the present invention can also be applied to synthesis of encoded image data created by an image encoding method such as a JPEG image encoding method, MPEG (Motion Picture Expert Group) 1, MPEG2 encoding method or the like.

  Furthermore, although the image synthesis apparatuses 1004 and 1504 in the first or second embodiment are described on the assumption that the quantization table is included in the encoded data, the quantization table is not included in the data. However, it only needs to be associated. In addition, the quantization table does not need to be in the form of a table in a strict sense, and may have a role for defining the quantization width of the function conversion coefficient.

The above-described image composition apparatus includes the following.
(1) Function transformation for transforming from a spatial domain into data represented by a function transformation coefficient value by a set of basis functions, one or more image data quantized and encoded, and the one or more images An input means for inputting a quantization table corresponding to each data and a synthetic quantization table;
Decoding means for decoding the plurality of input image data;
Extraction means for extracting a plurality of regions from the decoded one or more image data;
A synthesizing unit that generates synthetic data obtained by synthesizing the plurality of extracted regions;
Transform means for transforming the function transform coefficient value of each of the plurality of extracted regions based on a quantization table corresponding to the region and the combined quantization table;
Encoding means for encoding combined data in which the plurality of converted regions are combined;
An image synthesizing apparatus comprising output means for associating and outputting the encoded synthesized data and the synthesized quantization table.

(2) one or a plurality of image data and one or a plurality of image data that have been transformed, quantized and encoded by a function transform that transforms from a spatial domain into data represented by a function transform coefficient value by a set of basis functions Input means for inputting a quantization table corresponding to
Decoding means for decoding the plurality of input image data;
Extraction means for extracting a plurality of regions from the decoded one or more image data;
A synthesizing unit that generates synthetic data obtained by synthesizing the plurality of extracted regions;
Synthetic quantization table creation processing means for creating a synthetic quantization table based on at least the quantization tables corresponding to the plurality of extracted regions;
Transform means for transforming the function transform coefficient value of each of the plurality of extracted regions based on a quantization table corresponding to the region and the combined quantization table;
Encoding means for encoding combined data in which the plurality of converted regions are combined;
An image synthesizing apparatus comprising output means for associating and outputting the encoded synthesized data and the synthesized quantization table.

(3) one or a plurality of image data that has been transformed, quantized and encoded by a function transform that converts from a spatial domain into data represented by a function transform coefficient value by a set of basis functions, and the one or more image data An input means for inputting a quantization table corresponding to each and a synthetic quantization table;
Decoding means for decoding the plurality of input image data;
Extraction means for extracting a plurality of regions from the decoded one or more image data;
Inverse quantization means for inversely quantizing each of the extracted plurality of regions using a quantization table corresponding to the region;
A synthesizing unit that generates synthetic data obtained by synthesizing the plurality of extracted regions;
Quantization means for quantizing the synthesized data using the synthesized quantization table;
Encoding means for encoding the quantized composite data;
An image synthesizing apparatus comprising output means for associating and outputting the encoded synthesized data and the synthesized quantization table.

(4) One or a plurality of image data that has been transformed, quantized and encoded by function transformation for converting from a spatial domain to data represented by a function transformation coefficient value by a set of basis functions, and the one or more image data Input means for inputting a quantization table corresponding to each;
Decoding means for decoding the plurality of input image data;
Extraction means for extracting a plurality of regions from the decoded one or more image data;
Inverse quantization means for inversely quantizing each of the extracted plurality of regions using a quantization table corresponding to the region;
A synthesizing unit that generates synthetic data obtained by synthesizing the plurality of extracted regions;
A synthetic quantization table creation processing means for creating a synthetic quantization table based on a quantization table corresponding to each of the plurality of regions;
Quantization means for quantizing the synthesized data using the created synthesized quantization table;
Encoding means for encoding the quantized composite data;
An image synthesizing apparatus comprising output means for associating and outputting the encoded synthesized data and the synthesized quantization table.

  The embodiment disclosed this time should be considered as illustrative in all points and not restrictive. The scope of the present invention is defined by the terms of the claims, rather than the description above, and is intended to include any modifications within the scope and meaning equivalent to the terms of the claims.

It is a functional block diagram which shows the outline of the function of the image composition apparatus in the 1st Embodiment of this invention. It is a figure which shows an example of synthetic | combination map data. It is a figure which shows 1st image cutting area data. It is a figure which shows 2nd image cutting area data. It is a figure which shows synthetic image sticking area | region data. It is a figure for demonstrating another example of an image composition process. It is another figure which shows 1st image cutting area data. It is another figure which shows 2nd image cutting area data. It is another figure which shows synthetic image sticking area | region data. It is a figure which shows an example of a 1st quantization table. It is a figure which shows an example of a 2nd quantization table. It is a figure which shows an example of a synthetic | combination quantization table. It is a flowchart which shows the flow of the process performed with the image composition apparatus in 1st Embodiment. It is a functional block diagram which shows the outline of the function of the image synthesizing | combining apparatus in 2nd Embodiment. It is a flowchart which shows the flow of the image composition process performed with the image composition apparatus in 2nd Embodiment. It is a figure for demonstrating the synthetic | combination process of several image data.

Explanation of symbols

  1001, 1002 Image input device, 1003 Composition map input device, 1004, 1504 Image composition device, 1005 Coding composition image data output device, 1006, 1007 Data buffer, 1008 Composition map data buffer, 1009 Quantization function transform coefficient acquisition processing unit 1010 Quantization table acquisition processing unit, 1011, 1013 Quantization function conversion coefficient data buffer, 1012, 1015 Quantization table buffer, 1014 Synthesis quantization function coefficient data creation processing unit, 1016 Synthesis quantization function conversion coefficient data buffer, 1017 Synthesis quantization table creation processing unit, 1018 synthesis quantization table buffer, 1019 coded synthesized image data creation processing unit, 1020 coded synthesized image data buffer, 1508 synthesis map data buffer, 15 09 image decoding processing unit, 1511, 1513 image data buffer, 1514 synthesized image data creation processing unit, 1516 synthesized image data buffer, 1018 synthesized quantization table buffer, 1519 encoded image processing unit.

Claims (16)

One or a plurality of image data including a function transform and a quantized function transform coefficient value converted from the spatial domain into data represented by a function transform coefficient value by a set of basis functions, and each of the one or more image data Input means for inputting a quantization table corresponding to
Based on a plurality of quantization tables respectively corresponding to a plurality of regions of the input one or a plurality of image data, a synthesized quantization table creating processing means for creating a synthesized quantization table;
Transform means for transforming the plurality of regions based on the respective function transform coefficient values based on the quantization table corresponding to the region and the combined quantization table;
An image synthesizing apparatus, comprising: synthesized data generating means for generating synthesized data by synthesizing the plurality of converted areas.   2. The conversion unit according to claim 1, wherein the conversion unit converts the function conversion coefficient value of each of the plurality of regions based on a corresponding element of the quantization table of the region and a corresponding element of the combined quantization table. Image composition device. Corresponding to each of a plurality of first image data and a plurality of first image data including a function transform for converting from a spatial domain to data represented by a function transform coefficient value by a set of basis functions and a quantized function transform coefficient value First input means for inputting a quantization table to be
A second input means for inputting one or a plurality of second image data composed of spatial area image data defined in the spatial area;
Based on a plurality of quantization tables respectively corresponding to the plurality of first image data, a synthesized quantization table creating processing means for creating a synthesized quantization table;
Of the plurality of regions of the input first image data and second image data, function conversion coefficient values of the regions of the first image data are based on the quantization table corresponding to the region and the combined quantization table. Conversion processing means for converting spatial region image data corresponding to the region of the second image data out of the plurality of regions into a function conversion coefficient value by a set of basis functions,
An image synthesizing apparatus comprising: synthesized data generating means for further synthesizing the converted function transform coefficient value based on the synthesized quantization table and then synthesizing the plurality of regions to generate synthesized data.   The converting means converts the function conversion coefficient value of each of the plurality of regions from the first image data based on a corresponding element of the quantization table of the region and a corresponding element of the combined quantization table. The image composition device according to claim 3. Corresponding to each of a plurality of first image data and a plurality of first image data including a function transform for converting from a spatial domain to data represented by a function transform coefficient value by a set of basis functions and a quantized function transform coefficient value First input means for inputting a quantization table to be
A second input means for inputting one or a plurality of second image data composed of spatial area image data defined in the spatial area;
Decoding means for decoding each of the plurality of first image data into spatial domain image data;
Composition for generating spatial region image data obtained by decoding the first image data corresponding to a plurality of regions of the input first image data and second image data, and spatial region synthesized image data obtained by synthesizing the second image data. Means,
Based on a plurality of quantization tables corresponding to the plurality of first image data, a synthesized quantization table creating processing means for creating a synthesized quantization table;
An image synthesizing apparatus comprising: an encoding unit configured to generate the encoded synthesized image data by performing the function conversion on the spatial domain synthesized image data and then quantizing using the synthesized quantization table. A plurality of image data including a function transform for converting from a spatial domain to data represented by a function transform coefficient value by a set of basis functions and a quantized function transform coefficient value, and a quantization table corresponding to each of the plurality of image data An input means for inputting
Extracting means for extracting at least one region from each of the plurality of input image data;
Based on the plurality of quantization tables, a synthetic quantization table creation processing means for creating a synthetic quantization table;
For each extracted region, the function conversion coefficient value corresponding to each region is converted based on the quantization table corresponding to the image data from which the region is extracted and the combined quantization table, and then each converted region is converted. An image synthesizing apparatus comprising: synthesized data generating means for generating synthesized data by synthesizing region data. The composite data generation means converts the extracted function conversion coefficient value of each region based on a corresponding element of the quantization table of the region and a corresponding element of the composite quantization table. The image composition device described in 1. The combined quantized table creation processing means, each element of the combined quantized table, a common divisor between the corresponding elements in each of the plurality of quantization tables, according to any one of claims 1-7 Image composition device. The combined quantized table creation processing means, each element of the combined quantized table, the greatest common divisor of the corresponding element in each of said plurality of quantization tables, according to any one of claims 1-7 Image composition device. One or a plurality of image data including a function transform and a quantized function transform coefficient value converted from the spatial domain into data represented by a function transform coefficient value by a set of basis functions, and each of the one or more image data Inputting a quantization table corresponding to
Creating a composite quantization table based on a plurality of quantization tables respectively corresponding to a plurality of regions of the input one or more image data;
Transforming the plurality of regions based on respective function transform coefficient values based on a quantization table corresponding to the region and the combined quantization table;
Combining the plurality of converted regions to generate composite data. Corresponding to each of a plurality of first image data and a plurality of first image data including a function transform for converting from a spatial domain to data represented by a function transform coefficient value by a set of basis functions and a quantized function transform coefficient value Inputting a quantization table to be
Inputting one or a plurality of second image data composed of spatial region image data defined in the spatial region;
Creating a synthetic quantization table based on a plurality of quantization tables respectively corresponding to the plurality of first image data;
Of the plurality of regions of the input first image data and second image data, function conversion coefficient values of the regions of the first image data are based on the quantization table corresponding to the region and the combined quantization table. Converting the spatial area image data corresponding to the area of the second image data out of the plurality of areas into a function conversion coefficient value by a set of basis functions;
And further comprising: quantizing the converted function transform coefficient value based on the synthesized quantization table, and synthesizing the extracted plurality of regions to generate synthesized data. Corresponding to each of a plurality of first image data and a plurality of first image data including a function transform for converting from a spatial domain to data represented by a function transform coefficient value by a set of basis functions and a quantized function transform coefficient value Inputting a quantization table to be
Inputting one or a plurality of second image data composed of spatial region image data defined in the spatial region;
Decoding each of the plurality of first image data into spatial domain image data;
Composition for generating spatial region image data obtained by decoding the first image data corresponding to a plurality of regions of the input first image data and second image data, and spatial region synthesized image data obtained by synthesizing the second image data. Steps,
Creating a synthetic quantization table based on a plurality of quantization tables corresponding to the plurality of first image data;
And performing the function conversion on the spatial domain composite image data and then quantizing using the composite quantization table to generate encoded composite image data. A plurality of image data including a function transform for converting from a spatial domain to data represented by a function transform coefficient value by a set of basis functions and a quantized function transform coefficient value, and a quantization table corresponding to each of the plurality of image data An input step for entering
An extraction step of extracting a region from each of the plurality of input image data;
Based on the plurality of quantization tables, a synthetic quantization table creation processing step of creating a synthetic quantization table;
For each extracted region, after converting the function conversion coefficient value corresponding to each region based on the quantization table corresponding to the image data from which the region is extracted and the combined quantization table, An image synthesizing method, comprising: a synthesized data generation step of synthesizing region data to generate synthesized data. The combined data generation step converts the extracted function conversion coefficient value of each area based on a corresponding element of the quantization table of the area and a corresponding element of the combined quantization table. The image composition method described in 1. An image composition program for causing a computer to execute each step of the image composition method according to any one of claims 10 to 14 . A computer-readable recording medium on which the image composition program according to claim 15 is recorded.

Images