JP2018201143A - Picture encoder - Google Patents

Abstract

To provide a picture encoder capable of executing reversible compression and irreversible compression selectively, while making the circuit scale compact.SOLUTION: A picture encoder includes a frequency conversion part, a quantization part, a zigzag scan part, a coding part 6 and a first input changeover part, where the first input changeover part inputs image data while switching to the frequency conversion part and the coding part 6. The coding part 6 includes a distribution part 8 for outputting while dividing into DC coefficients and AC coefficients, AC component processing parts 10, 11 for encoding the AC coefficients inputted from the distribution part 8, a second input changeover part 9 for outputting the DC coefficients inputted from the distribution part 8, and image data inputted externally while switching, DC component processing parts 12, 13 for calculating the differential value of data outputted from the second input changeover part 9 before and after and encoding, and a code junction part 14 for sequentially joining the code data outputted from the DC component processing parts 12, 13, and outputting the code data, outputted from the DC component processing parts 12, 13 and the AC component processing parts 10, 11, while joining.SELECTED DRAWING: Figure 2

Description

  The present invention relates to an image coding apparatus that compresses image data using a Huffman code, and more particularly to an image coding apparatus that can compress image data by switching between lossless compression and lossy compression.

  2. Description of the Related Art Conventionally, in the field of handling images such as photographs, an image encoding device that encodes and compresses image data as original data according to the JPEG (Joint Photographic Experts Group) standard has been used for the purpose of reducing the burden on communication. ing.

  As such an image compression method, conventionally, an irreversible compression method (lossy) in which a high compression rate is obtained although the original data is lost by encoding, and a lossless compression method (lossless) in which the original data is not lost although the compression rate is low. The lossy compression method is used for general image compression, and the lossless compression method is used in fields where high image quality is required.

  By the way, the lossless compression type image encoding device and the lossy compression type image encoding device are generally configured as separate devices, but high compression rate image compression and high image quality image compression. Depending on the situation, there are cases where you want to use them properly. In order to enable such use, conventionally, a high-compression image compression function (lossy compression function) and a high-quality image compression function (reversible compression function) There has been proposed an image encoding apparatus in which the two functions are incorporated into one apparatus (see also Japanese Unexamined Patent Application Publication No. 2009-105700 below).

  FIG. 3 and FIG. 4 show a general configuration of an image encoding apparatus having such two functions of a lossy compression function and a lossless compression function. As shown in FIG. 3, the image encoding device 100 includes an irreversible compression unit 110, a lossless compression unit 130, an input switching unit 101, and an output switching unit 102.

  The lossy compression unit 110 includes a DCT unit 111, a quantization unit 112, a zigzag scanning unit 113, and a Huffman encoding unit 114, and the lossless compression unit 130 includes a Huffman encoding unit 131.

  The DCT unit 111 processes image data converted from RGB color data to YUV luminance color difference data in units of 8 × 8 pixel blocks, and converts the image data of each pixel in block units to frequency components (DCT (Discrete)). (Referred to as a Cosine Transform coefficient) to generate coefficient data composed of one DC (Direct Current) coefficient and 63 AC (Alternating Current) coefficients.

  In addition, the quantization unit 112 converts each 8 × 8 coefficient data (numerical value) generated by the DCT unit 111 in units of blocks into a quantization table (a table having 8 × 8 numerical values) prepared in advance. ) Quantization is performed by dividing by each corresponding numerical value. In addition, the zigzag scanning unit 113 linearly arranges the 8 × 8 coefficient data quantized in block units in the quantization unit 112 by scanning in a zigzag manner (that is, arranges it in one dimension). The Huffman encoder 114 performs a process of generating code data by encoding the coefficient data in units of blocks arranged one-dimensionally by the zigzag scan unit 113 using a Huffman code.

  As shown in FIG. 4, the Huffman encoding unit 114 includes a marker storage unit 115, a distributor 116, a DC difference calculation unit 117, a DC component encoding unit 118, a zero-run counter unit 119, an AC component encoding unit 120, and a code. The connecting part 121 is configured.

  The marker storage unit 115 is a functional unit that stores markers, which will be described in detail later, and stores a DHT (Define Huffman Table) marker that defines a Huffman code table as one of the markers. Further, the distributor 116 outputs DC coefficient data to the DC difference calculation unit 117 among the coefficient data output from the zigzag scanning unit 113 in units of blocks, and outputs AC coefficient data to the zero-run counter unit 119. Perform the process.

  The DC difference calculation unit 117 temporarily stores sequentially input DC coefficient data, and calculates and outputs a difference between the input current DC coefficient data and the previous DC coefficient data input immediately before. The DC component encoding unit 118 processes the difference value output from the DC difference calculation unit 117 using the Huffman encoding table for DC component stored in the marker storage unit 115, and obtains group number information. A process of generating and outputting variable length code data in units of the encoded Huffman code and the subsequent additional bits related to the difference value is performed.

  On the other hand, the zero-run counter unit 119 counts the number of consecutive AC coefficient data having a value of zero (zero-run information) when the value of the sequentially input AC coefficient data is zero, and determines the AC component code. In the case of AC coefficient data that is output to the encoding unit 120 and whose value is not zero, the processing unit outputs the AC coefficient data to the AC component encoding unit 120 as it is. Further, the AC component encoding unit 120 processes the data output from the zero-run counter unit 119 using the Huffman encoding table for the AC component stored in the marker storage unit 115, and the zero-run information and the group Processing for generating and outputting AC component variable length code data in units of additional bits related to the Huffman code obtained by encoding the number information and the AC coefficient value following the Huffman code is performed.

  The code concatenation unit 121 includes various markers stored in the marker storage unit 115, variable length code data output from the DC component encoding unit 118, and variable output from the AC component encoding unit 120. The long code data is concatenated and output as one code data.

  FIG. 5A shows the data structure of the code data (JPEG stream) generated by the lossy compression method as described above. As shown in FIG. 5A, this code data includes a front marker group composed of various markers, encoded compressed image data, and an EOI (End Of Image) marker that defines the end of the code data. Are sequentially arranged.

  The front marker group defines an SOI (Start Of Image) marker that defines the beginning of code data, a DQT (Define Quantization Table) marker that defines the quantization table, a frame start, and an image size. The SOF (Start Of Frame) marker, the DHT marker that defines the Huffman code table, and the SOS (Start Of Scan) marker that defines the presence of compressed image data thereafter are included. The compressed image data has a structure in which the variable length code data in units of blocks are continuously connected.

  On the other hand, the Huffman encoding unit 131 includes a marker storage unit 132, a difference calculation unit 133, an encoding unit 134, and a code connecting unit 135. The marker storage unit 132 is a functional unit that stores various markers in the same manner as the marker storage unit 115, and stores a DHT marker that defines a Huffman code table as one of the stored markers.

  The difference calculating unit 133 temporarily stores sequentially input image data, performs a process of calculating and outputting a difference between the input current image data and the previous image data input immediately before, and performs encoding. The unit 134 processes the difference value output from the difference calculation unit 133 using the Huffman coding table for DC component stored in the marker storage unit 132, and encodes the group number information, and Subsequent processing for generating and outputting variable-length code data in units of additional bits related to the difference value is performed.

  The code concatenation unit 135 performs a process of concatenating various markers stored in the marker storage unit 132 and variable-length code data output from the encoding unit 134 and outputting the data as one code data.

  FIG. 5B shows the data structure of the code data (JPEG stream) generated in this way by the lossless compression method. As shown in FIG. 5B, this code data includes a front marker group composed of various markers, encoded compressed image data, and an EOI (End Of Image) marker that defines the end of the code data. Are sequentially arranged. The front marker group of code data in the lossless compression method is different from the front marker group of code data in the lossy compression method in that it does not have a DQT marker.

  The input switching unit 101 switches the image data to the irreversible compression unit 110 and the reversible compression unit 130 in accordance with a switching signal input from outside. Specifically, when a signal requesting lossy compression is received, the input switching unit 101 inputs image data to the lossy compression unit 110, and when receiving a signal requesting lossless compression, the input switching unit 101 receives image data. Is input to the lossless compression unit 130.

  Similarly, the output switching unit 102 switches between the code data generated by the lossy compression unit 110 and the code data generated by the lossless compression unit 130 according to a switching signal input from outside and outputs the code data to the outside. Specifically, when the output switching unit 102 receives a signal requesting lossy compression, the output switching unit 102 outputs the code data generated by the lossy compression unit 110, and when receiving a signal requesting lossless compression, the lossless compression unit The code data generated by 130 is output.

  According to the image coding apparatus 100 having the above configuration, when a signal requesting irreversible compression is input, image data is input to the irreversible compression unit 110 via the input switching unit 101, and Code data generated by the compression unit 110 is output to the outside via the output switching unit 102. On the other hand, when a signal requesting lossless compression is input, the image data is input to the lossless compression unit 130 via the input switching unit 101, and the code data generated by the lossless compression unit 130 is output via the output switching unit 102. Output to the outside. Thus, according to the image encoding device 100, the image data can be selectively compressed by either the lossless compression method or the lossy compression method.

JP 2009-105700 A

  By the way, the above-described conventional image encoding device 100 simply includes the irreversible compression unit 110 and the lossless compression unit 130 arranged in parallel, and when the image encoding device 100 is constructed by an electronic circuit, Since electronic circuits corresponding to irreversible compression and reversible compression are required, the circuit scale is not much different from the case where they are configured as separate devices, i.e., independent electronic circuits. There was no merit in terms of cost.

  In recent years, there has been a continuous search for electronic circuit miniaturization, and image coding apparatuses are also required to have a smaller circuit scale. And the manufacturing cost of an apparatus can be made low by making a circuit scale small.

  The present invention has been made in view of the above circumstances, and is capable of selectively executing lossless compression and lossy compression, and can further reduce the circuit scale as compared with the prior art. The purpose is to provide a device.

The present invention for solving the above problems is as follows.
A frequency conversion unit that generates image coefficient data and AC coefficient data by converting image data into coefficients by frequency conversion in units of a predetermined number of pixels in double rows and multiple columns;
A quantization unit that quantizes each of the block-unit DC coefficient data and AC coefficient data generated by the frequency conversion unit;
A zigzag scanning unit that zigzags the block unit DC coefficient data and the AC coefficient data quantized by the quantization unit to arrange them in a line;
In an image encoding device including an encoding unit that encodes the DC coefficient data and the AC coefficient data in units of blocks arranged by the zigzag scanning unit,
A first input switching unit for switching and inputting the image data to the frequency conversion unit and the encoding unit;
The encoding unit includes:
The input AC coefficient data is processed using the AC component Huffman coding table, and the Huffman code obtained by encoding the zero-run information and the group number information, and the subsequent AC bits in units of additional bits related to the AC coefficient value. An AC component processing unit that generates component variable-length code data;
The difference value between the input current data and the previous data input immediately before is calculated, the calculated difference value is processed using a Huffman encoding table for DC components, and group number information is encoded A DC component processing unit that generates variable length code data in units of additional Huffman codes and subsequent additional bits related to the difference value;
When variable length code data is output only from the DC component processing unit connected to the DC component processing unit and the AC component processing unit, the output variable length code data is sequentially connected and output, and When variable length code data is output from the DC component processing unit and the AC component processing unit, the variable length code data output from the DC component processing unit and the AC component processing unit are sequentially connected and output. A sign concatenation,
A data distribution unit that is connected to the zigzag scanning unit and outputs the coefficient data output from the zigzag scanning unit separately into DC coefficient data and AC coefficient data, and outputs the AC coefficient data to the AC component processing unit When,
A second input switching unit that is connected to the data distribution unit and switches between the DC coefficient data output from the data distribution unit and the image data input from the first input switching unit, and is input to the DC component processing unit; The present invention relates to an image encoding device comprising:

  According to this image encoding apparatus, when a signal requesting lossy compression is input from the outside, the first input switching unit receives this request signal and inputs the input image data to the frequency conversion unit. Perform the process.

  In the frequency conversion unit, the input image data is converted into coefficients by frequency conversion in units of blocks to generate DC coefficient data and AC coefficient data, and then the generated DC coefficient data and AC coefficient data are The block unit DC coefficient data and the AC coefficient data quantized by the quantization unit are arranged in a line by the zigzag scanning unit and sequentially input to the data distribution unit of the encoding unit.

  The data distribution unit outputs DC coefficient data among the coefficient data output from the zigzag scanning unit to the second input switching unit, and outputs AC coefficient data to the AC component processing unit. The second input switching unit receives a signal requesting irreversible compression from the outside, and receives the DC coefficient data input from the data distribution unit to the DC component processing unit. The DC component processing unit calculates a difference value between the current DC coefficient data that is sequentially input and the previous DC coefficient data that is input immediately before it, and the calculated difference value is used for the DC component. To generate variable length code data in units of additional bits related to the difference value, and a Huffman code obtained by encoding the group number information, and then to the code concatenation unit. Output.

  On the other hand, the AC component encoding unit processes sequentially input AC coefficient data using an AC component Huffman encoding table, encodes the zero-run information and the group number information, and the subsequent AC coefficients. AC component variable-length code data is generated in units of additional bits related to numerical values, and input to the code concatenation unit.

  Then, the variable length code data sequentially output from the DC component processing unit and the AC component processing unit are concatenated by the code concatenating unit and output to the outside as code data compressed by the lossy compression method.

  On the other hand, when a signal requesting lossless compression is input from the outside, the first input switching unit receives this request signal and inputs the input image data to the second input switching unit of the encoding unit. . The second input switching unit receives the signal requesting the lossless compression from the outside, and receives the signal to input the image data sequentially input from the first input switching unit to the DC component processing unit. The DC component processing unit calculates a difference value between the current image data sequentially input and the previous image data input immediately before it, and the calculated difference value is used for the DC component. Process using the Huffman coding table to generate a Huffman code obtained by coding the group number information, and variable length code data in units of additional bits related to the difference value, and output to the code concatenation unit To do.

  The variable-length code data sequentially output from the DC component processing unit is concatenated by the code concatenation unit and output to the outside as code data compressed by the lossless compression method.

  As described above, according to the image coding apparatus according to the present invention, by inputting a signal requesting lossless compression or lossy compression from the outside, code data (compression data) obtained by compressing image data in accordance with the input request. Image data) is generated and output.

  In this image encoding device, lossless compression and lossy compression are processed in the same encoding unit. Therefore, when the image encoding device is constructed by an electronic circuit, lossless compression and lossy compression are performed. Compared to a conventional image encoding apparatus configured with separate electronic circuits, the circuit scale can be reduced, and the manufacturing cost can be reduced.

  As described above, according to the image coding apparatus according to the present invention, by inputting a request signal for lossless compression or lossy compression, code data obtained by compressing image data in response to the request is generated. Is output.

  In this image encoding device, lossless compression and lossy compression are processed in the same encoding unit. Therefore, when the image encoding device is constructed by an electronic circuit, lossless compression and lossy compression are performed. Compared to a conventional image encoding apparatus configured with separate electronic circuits, the circuit scale can be reduced, and the manufacturing cost can be reduced.

It is the block diagram which showed schematic structure of the image coding apparatus which concerns on one Embodiment of this invention. It is the block diagram which showed the structure of the Huffman encoding part of the image coding apparatus which concerns on this embodiment. It is the block diagram which showed schematic structure of the conventional image coding apparatus. It is the block diagram which showed the structure of the Huffman encoding part in the conventional image coding apparatus. It is explanatory drawing which shows the data structure of the compressed code data, (a) is explanatory drawing which showed the data structure of the code data compressed by the lossy compression system, (b) is the data of the code data compressed by the lossless compression system It is explanatory drawing which showed the structure.

  Hereinafter, an embodiment of the present invention will be described with reference to the drawings. As shown in FIG. 1, the image encoding device 1 includes a first input switching unit 2, a DCT unit 3, a quantization unit 4, a zigzag scanning unit 5, and a Huffman encoding unit 6. The image encoding device 1 can be realized by an electronic device (semiconductor integrated circuit or the like) provided with an electronic circuit as appropriate.

  When the first input switching unit 2 receives a signal requesting reversible compression selectively input from the outside and a signal requesting irreversible compression, and receiving a signal requesting irreversible compression, When the sequentially input image data is input to the DCT unit 3 and a signal requesting lossless compression is received, a process of inputting the sequentially input image data to the second input switching unit 9 described later in detail. Do.

  Similar to the conventional DCT unit 111 described above, the DCT unit 3 processes image data converted from RGB color data to YUV luminance color difference data in units of 8 × 8 pixels, and each pixel in units of blocks. Each of the image data is frequency-converted to generate coefficient data composed of one DC coefficient and 63 AC coefficients.

  Also, the quantization unit 4, similar to the above-described conventional quantization unit 112, uses 8 × 8 coefficient data (numerical values) generated by the DCT unit 3 in units of blocks, The quantization process is performed by dividing each of the corresponding numerical values in the quantization table (table having 8 × 8 numerical values).

  Similarly to the conventional zigzag scanning unit 5 described above, the zigzag scanning unit 5 linearly arranges 8 × 8 coefficient data quantized in units of blocks in a zigzag manner (that is, 1 Process).

  As shown in FIG. 2, the Huffman encoding unit 6 includes a marker storage unit 7, a distributor 8, a zero-run counter unit 10, an AC component encoding unit 11, a difference calculation unit 12, a DC component encoding unit 13, and a code. It is composed of a connecting part 14.

  The marker storage unit 7 is a functional unit that stores various markers in the same manner as the conventional marker storage unit 115 described above. Specifically, the marker storage unit 7 includes a front marker group of code data. The SOI marker, DQT marker, SOF marker, DHT marker, SOS marker, and EOI marker that defines the end of the code data are stored.

  The distributor 8 outputs DC coefficient data to the second input switching unit 9 among the coefficient data output from the zigzag scanning unit 5 in units of blocks, and outputs AC coefficient data to the zero-run counter unit 10. Perform the process. Similarly to the first input switching unit 2, the second input switching unit 9 receives a signal requesting reversible compression selectively input from the outside and a signal requesting irreversible compression. When a signal requesting lossless compression is received, DC coefficient data sequentially input from the distributor 8 is input to the difference calculation unit 12, and when a signal requesting lossless compression is received, the first A process of inputting image data sequentially input from the 1-input switching unit 2 to the difference calculation unit 12 is performed.

  When the value of the AC coefficient data sequentially input from the distributor 8 is zero, the zero run counter unit 10 counts the number (zero run information) of consecutive AC coefficient data having a value of zero. In the case of AC coefficient data that is output to the AC component encoding unit 11 and whose value is not zero, a process of outputting to the AC component encoding unit 11 as it is is performed.

  Then, the AC component encoding unit 11 processes the data output from the zero run counter unit 10 using the Huffman encoding table for the AC component stored in the marker storage unit 7 to obtain the zero run information and the group Processing for generating and outputting AC component variable length code data in units of additional bits related to the Huffman code obtained by encoding the number information and the AC coefficient value following the Huffman code is performed.

  On the other hand, the difference calculation unit 12 is a functional unit that processes DC coefficient data or image data sequentially output from the second input switching unit 9, and inputs data sequentially output from the second input switching unit 9. Then, a process of temporarily holding and calculating and outputting a difference between the input current data and the previous data input immediately before is performed.

  Further, the DC component encoding unit 13 processes the difference value output from the difference calculation unit 12 using the Huffman encoding table for DC component stored in the marker storage unit 7 to obtain group number information. A process of generating and outputting variable length code data in units of the encoded Huffman code and the subsequent additional bits related to the difference value is performed. Note that the same table can be used as the Huffman encoding table for encoding the difference value of the DC coefficient data in the lossy compression and the Huffman encoding table for encoding the difference value of the image data in the lossless compression. The same table is used in the example.

  When the variable-length code data is output only from the DC component encoding unit 13, that is, in the case of lossless compression, the code connecting unit 14 is a reversible among the various markers stored in the marker storage unit 7. The compression marker is read out, and the read marker and the variable length code data sequentially output from the DC component encoding unit 13 are concatenated and output as one code data.

  On the other hand, when variable-length code data is output from the DC component encoding unit 13 and the AC component encoding unit 11, that is, in the case of lossy compression, the code concatenating unit 14 includes the marker storage unit 7 Among the various markers stored in, the lossy compression marker is read out, the read marker, the variable length code data output from the DC component encoding unit 13 and the AC component encoding unit 11 are output. The variable length code data is concatenated and output as one code data.

  As described above, the lossless compression marker and the lossy compression marker are the lossy compression front marker group in that the lossless compression front marker group does not have a DQT marker. The code data by reversible compression has the data structure shown in FIG. 5B, and the code data by lossy compression has the data structure shown in FIG. 5A.

  According to the image encoding device 1 of the present example having the above configuration, when a signal requesting lossy compression is input from the outside, the first input switching unit 2 receives this request signal, The input image data is input to the DCT unit 3. In the DCT unit 3, the input image data is coefficientized by frequency conversion in units of blocks to generate DC coefficient data and AC coefficient data. The generated DC coefficient data and AC coefficient data are respectively quantized by the quantizing unit 4, and the quantized DC coefficient data and AC coefficient data in units of blocks are arranged in a line by the zigzag scanning unit 5, The signals are sequentially input to the distributor 8 of the Huffman encoder 6.

  The distributor 8 outputs DC coefficient data among the coefficient data output from the zigzag scanning unit 5 to the second input switching unit 9 and outputs AC coefficient data to the zero-run counter unit 10. Then, the second input switching unit 9 receives the request signal related to irreversible compression, outputs the DC coefficient data input from the distributor 8 to the difference calculation unit 12, and the difference calculation unit 12 Calculates the difference between the DC coefficient data before and after sequentially input and outputs the difference to the DC component encoding unit 13, and the DC component encoding unit 13 stores the input difference value in the DC stored in the marker storage unit 7. Processing using the component Huffman coding table to generate a Huffman code obtained by coding group number information and variable length code data in units of additional bits related to the difference value following the Huffman code. Input to section 14.

  On the other hand, when the value of the AC coefficient data sequentially input from the distributor 8 is zero, the zero run counter unit 10 counts the number of consecutive AC coefficient data having a value of zero (zero run information). In the case of AC coefficient data having a value that is not zero, the AC component encoding unit 11 performs a process of outputting the AC component data as it is to the AC component encoding unit 11. The data output from the run counter unit 10 is processed using the AC component Huffman encoding table stored in the marker storage unit 7 to encode the zero run information and the group number information, and to this The variable length code data of the AC component in units of additional bits related to the subsequent AC coefficient value is generated and input to the code concatenation unit 14.

  In the code concatenation unit 14, the lossy compression markers stored in the marker storage unit 7 and the variable length code data sequentially output from the DC component encoding unit 13 and the AC component encoding unit 11 are concatenated. The data is output to the outside as code data compressed by a lossy compression method.

  On the other hand, when a signal requesting reversible compression is input from the outside, the first input switching unit 2 receives the request signal, and sequentially inputs image data to the second Huffman encoding unit 6. Input to the input switching unit 9. When the second input switching unit 9 receives the request signal related to the lossless compression, the second input switching unit 9 inputs the image data sequentially input from the first input switching unit 2 to the difference calculating unit 12, and the difference calculating unit 12 The difference between the image data before and after sequentially input is calculated and output to the DC component encoding unit 13, and the DC component encoding unit 13 uses the input difference value for the DC component stored in the marker storage unit 7. Processing using the Huffman coding table generates Huffman code obtained by coding the group number information, and variable length code data in units of additional bits related to the difference value following the Huffman code. input.

  In the code concatenation unit 14, the lossless compression marker stored in the marker storage unit 7 and the variable length code data sequentially output from the DC component encoding unit 13 are concatenated and compressed by the lossless compression method. It is output to the outside as code data.

  As described above, according to the image encoding device 1 of the present example, by inputting a signal requesting lossless compression or lossy compression from the outside, code data (compressed) that compresses image data in accordance with the input request. Image data) is generated and output.

  Further, in this image encoding device 1, the lossless compression and the lossy compression are processed by the same Huffman encoding unit 6, so that when the image encoding device 1 is constructed by an electronic circuit, the lossless compression is performed. Compared with the case where the lossy compression and the lossy compression are configured by separate electronic circuits, the circuit scale can be reduced, and the manufacturing cost can be reduced.

  As mentioned above, although one Embodiment of this invention was described, the specific aspect which this invention can take is not limited to this at all.

  For example, the above-described image encoding device 1 can be configured from a computer including a CPU, a RAM, a ROM, and the like. The first input switching unit 2, the DCT unit 3, the quantization unit 4, and the zigzag scanning unit 5. The function can be realized by a computer program as appropriate. Further, the distributor 8, the second input switching unit 9, the zero run counter unit 10, the AC component encoding unit 11, the difference calculation unit 12, the DC component encoding unit 13 and the code concatenation unit 14 in the Huffman encoding unit 6 are also provided. The function can be realized by an appropriate computer program, and the marker storage unit 7 can be configured from an appropriate storage medium such as a RAM.

DESCRIPTION OF SYMBOLS 1 Image coding apparatus 2 1st input switching part 3 DCT part 4 Quantization part 5 Zigzag scanning part 6 Huffman encoding part 7 Marker memory | storage part 8 Distributor 9 2nd input switching part 10 Zero run counter part 11 AC component encoding Unit 12 difference calculation unit 13 DC component encoding unit 14 code concatenation unit

Claims (1)

A frequency conversion unit that generates image coefficient data and AC coefficient data by converting image data into coefficients by frequency conversion in units of a predetermined number of pixels in double rows and multiple columns;
A quantization unit that quantizes each of the block-unit DC coefficient data and AC coefficient data generated by the frequency conversion unit;
A zigzag scanning unit that zigzags the block unit DC coefficient data and the AC coefficient data quantized by the quantization unit to arrange them in a line;
In an image encoding device including an encoding unit that encodes the DC coefficient data and the AC coefficient data in units of blocks arranged by the zigzag scanning unit,
A first input switching unit for switching and inputting the image data to the frequency conversion unit and the encoding unit;
The encoding unit includes:
The input AC coefficient data is processed using the AC component Huffman coding table, and the Huffman code obtained by encoding the zero-run information and the group number information, and the subsequent AC bits in units of additional bits related to the AC coefficient value. An AC component processing unit that generates component variable-length code data;
The difference value between the input current data and the previous data input immediately before is calculated, the calculated difference value is processed using a Huffman encoding table for DC components, and group number information is encoded A DC component processing unit that generates variable length code data in units of additional Huffman codes and subsequent additional bits related to the difference value;
When variable length code data is output only from the DC component processing unit connected to the DC component processing unit and the AC component processing unit, the output variable length code data is sequentially connected and output, and When variable length code data is output from the DC component processing unit and the AC component processing unit, the variable length code data output from the DC component processing unit and the AC component processing unit are sequentially connected and output. A sign concatenation,
A data distribution unit that is connected to the zigzag scanning unit and outputs the coefficient data output from the zigzag scanning unit separately into DC coefficient data and AC coefficient data, and outputs the AC coefficient data to the AC component processing unit When,
A second input switching unit that is connected to the data distribution unit and switches between the DC coefficient data output from the data distribution unit and the image data input from the first input switching unit, and is input to the DC component processing unit; An image encoding device comprising:

Images