JP2009177353A - Encoder, method of controlling the same, and computer program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an encoder for suppressing deterioration in image quality in the boundary of an object by simple processing and a small bus rate. <P>SOLUTION: The encoder divides an image into a plurality of macroblocks for encoding. The encoder has: a generation means for generating a setting instruction for designating a quantization parameter for each of the plurality of macroblocks; a decision means for deciding the quantization parameters of macroblocks based on the setting instruction generated by the generation means; a conversion means for performing orthogonal transformation of the macroblocks and quantization using the decided quantization parameters; and an encoding means for subjecting the result of the conversion means to variable length encoding. The encoder calculates dispersion for a subblock composing the macroblocks, and generates the setting instruction of the quantization parameters based on the dispersion of the subblocks composing the macroblocks and that of the subblocks adjacent to the macroblocks. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to an encoding apparatus, an encoding apparatus control method, and a computer program.

  With the recent development of multimedia, various video compression encoding methods have been proposed. Typical examples are MPEG-1, 2, 4 and H.264. There is something like H.264. In these compression encoding processes, an original image (image) included in a moving image is divided into predetermined regions called blocks, and motion compensation prediction and DCT conversion processing are performed in units of the divided blocks. . In addition, when performing motion compensation prediction, an image obtained by local decoding of already encoded image data is used as a reference image, so that a decoding process is required even when encoding is performed.

  In addition, when compressing and encoding an image in accordance with the MPEG system, the amount of code often varies greatly depending on the spatial frequency characteristics, which are characteristics of the image itself, the scene, and the quantization scale value. An important technique for realizing a decoded image with good image quality in realizing an encoding apparatus having such encoding characteristics is code amount control.

  TM5 (Test Model 5) is generally used as one of the code amount control algorithms. The code amount control algorithm according to TM5 is composed of the following three steps, and the code amount is controlled in the following three steps so that the bit rate is constant for each GOP (Group Of Picture).

(STEP1)
The target code amount of the picture to be encoded from now is determined. Rgop, which is a code amount that can be used in the current GOP, is calculated by the following equation (1).
Rgop = (ni + np + nb) * (bits_rate / picture_rate) (1)
Here, ni, np, and nb are the number of remaining pictures in the current GOP of I, P, and B pictures, bits_rate represents the target bit rate, and picture_rate represents the picture rate.

Further, the complexity of the picture is obtained from the encoding result for each of the I, P, and B pictures by the following equation (2).
Xi = Ri * Qi
Xp = Rp * Qp (2)
Xb = Rb * Qb
Here, Xi, Xp, and Xb are also called complexity. Ri, Rp, and Rb are code amounts obtained as a result of encoding I, P, and B pictures, respectively. Further, Qi, Qp and Qb are average values of Q scales in all macroblocks in the I, P and B pictures, respectively. From the equations (1) and (2), the target code amounts Ti, Tp, and Tb for each of the I, P, and B pictures can be obtained by the following equation (3). Ti = max {(Rgop / (1 + ((Np * Xp) / (Xi * Kp)) + ((Nb * Xb) / (Xi * Kb)))), (bit_rate / (8 * picture_rate))}
Tp = max {(Rgop / (Np + (Nb * Kp * Xb) / (Kb * Xp))), (bit_rate / (8 * picture_rate))}
Tb = max {(Rgop / (Nb + (Np * Kb * Xp) / (Kp * Xb))), (bit_rate / (8 * picture_rate))}
... (3)
Np and Nb are the remaining number of P and B pictures in the current GOP, respectively, and constants Kp = 1.0 and Kb = 1.4.

(STEP2)
Three virtual buffers are used for each of the I, P, and B pictures, and the difference between the target code amount and the generated code amount obtained by Expression (3) is managed. The data accumulation amount of the virtual buffer is fed back, and the reference value of the Q scale is set for the macroblock to be encoded next so that the actual generated code amount approaches the target code amount based on the data accumulation amount. For example, when the current picture type is a P picture, the difference between the target code amount and the generated code amount can be obtained by arithmetic processing according to the following equation (4).

dp, j = dp, 0 + Bp, j−1 − ((Tp * (j−1)) / MB_cnt) (4)
Here, the subscript j is the number of the macroblock in the picture, dp, 0 indicates the initial fullness of the virtual buffer, Bp, j is the total code amount up to the jth macroblock, and MB_cnt is the macroblock in the picture Indicates a number. Next, when the reference value of the Q scale in the j-th macroblock is obtained using dp, j (hereinafter referred to as “dj”), equation (5) is obtained.

Qj = (dj * 31) / r (5)
Where r = 2 * bits_rate / picture_rate (6)
It is.

(STEP3)
A process of finally determining the quantization scale is executed based on the spatial activity of the macroblock to be encoded so that the visual characteristics, that is, the image quality of the decoded image is improved.

ACTj = 1+ min (vblk1, vblk2, ..., vblk8) (7)
In the equation (7), vblk1 to vblk4 indicate spatial activities in 8 × 8 sub-blocks in the macroblock of the frame structure. Further, vblk5 to vblk8 indicate spatial activities of 8 × 8 sub-blocks in the field-structure macroblock. Here, the calculation of the space activity can be obtained by the following equations (8) and (9).
vblk = Σ (Pi−Pbar) ² (8)
Pbar = (1/64) * ΣPi (9)
Here, Pi is a pixel value in the i-th macroblock, and Σ in equations (8) and (9) is an operation of i = 1 to 64. Next, ACTj obtained by the equation (7) is normalized by the following equation (10).

N_ACTj = (2 * ACTj + AVG_ACT) / (ACTj + AVG_ACT) (10)
Here, AVG_ACT is a reference value of ACTj in a previously encoded picture, and finally a quantization scale (Q scale value) MQUANTj is obtained by the following equation (11).

MQUANTj = Qj * N_ACTj (11)
According to the above TM5 algorithm, a large amount of code is allocated to the I picture by the processing of STEP1, and the code amount is in a flat portion (low spatial activity) that is visually noticeable in the picture. A lot will be allocated. Therefore, it is possible to perform code amount control and quantization control while suppressing deterioration in image quality within a predetermined bit rate.

Also, other techniques for performing quantization control according to image characteristics as in TM5 have been proposed, and visual improvement can be realized (see, for example, Patent Document 1).
JP 2005-323312 A

  The TM5 method described above performs quantization control so as to obtain a predetermined target code amount by extracting features in units of macroblocks and changing quantization parameters based on the features.

  However, since images with high dispersion include areas that are visually noticeable to humans, such as object boundaries, if the quantization is coarsened for such areas, visual degradation will occur at the object boundary. There is a problem that occurs.

  The method in Patent Document 1 also presents a solution to the same problem, but there is a problem that the processing amount and the bus rate increase because the quantization parameter is associated prior to encoding. .

  The present invention has been made in view of the above-described problems, and an object of the present invention is to provide an encoding apparatus that can suppress image quality deterioration at an object boundary with simple processing and a low bus rate.

The present invention for solving the above-described problem is an encoding apparatus that performs encoding by dividing an image into a plurality of macroblocks,
Generating means for generating a setting instruction for specifying a quantization parameter used for quantization for each of the plurality of macroblocks;
Determining means for determining a quantization parameter of the macroblock based on the setting instruction generated by the generating means;
Transform means for performing orthogonal transformation of the macroblock and quantization using the determined quantization parameter;
Encoding means for variable-length encoding the conversion result by the conversion means;
With
The generating means calculates a variance for subblocks constituting the macroblock, and based on the variance of subblocks constituting the macroblock and the variance of subblocks adjacent to the macroblock, the quantization parameter A setting instruction is generated.

  According to the present invention, it is possible to provide an encoding apparatus capable of suppressing image quality deterioration at an object boundary with simple processing and a low bus rate.

  Embodiments of the invention will be described below with reference to the accompanying drawings.

[First Embodiment]
A first embodiment will be described with reference to FIGS. FIG. 1 is an overall system configuration diagram in the first embodiment, and FIG. 2 is a diagram for explaining feature extraction.

  In FIG. 1, an input signal 101 is an input signal to an encoding device, and is input in a state of being divided into a plurality of predetermined blocks. The block is, for example, a block such as 16 × 16 or 8 × 8 in MPEG, and the size is determined according to the encoding method. In the following description, it will be called a macroblock.

  The encoding apparatus also includes an addition / subtraction unit 102, an orthogonal transform (DCT) unit 103, a quantization unit 104, an inverse quantization unit 105, an inverse orthogonal transform (inverse DCT) unit 106, a motion estimation unit 107, a motion compensation unit 108, A frame memory 109 and a variable length encoding unit 110 are provided. Further, a quantization control unit 112 and a feature extraction unit 113 are provided, and an output signal 116 is output.

  The operation of this encoding apparatus will be described. The input signal 101 divided into a plurality of blocks is subjected to orthogonal transform processing on the signals in the macroblock in the orthogonal transform unit 103 when the coded picture is an intra-frame coded (intra coded) picture. Then, the quantization unit 104 quantizes the DCT coefficient as the transformation result in the orthogonal transformation unit 103.

  When the encoded picture is an inter-frame encoded picture (inter-coded) picture, the inverse quantization unit 105 performs inverse quantization on the encoded picture, and the inverse orthogonal transform unit 106 performs inverse orthogonal transform processing. The performed local decoded image is created. Then, motion estimation with the coding target picture is performed by the motion estimation unit 107, motion compensation is performed by the motion compensation unit 108, and a difference value from the local decoded image is calculated by the addition / subtraction unit 102. The difference value is subjected to orthogonal transformation processing on the signal in the macroblock in the orthogonal transformation unit 103, and the quantization unit 104 quantizes the DCT coefficient.

  Regardless of intra-frame coding or inter-frame coding, the quantized signal quantized by the quantizing unit 104 is encoded by the variable-length encoding unit 110, and the encoded signal is output as the output signal 116. .

  Next, the code amount control unit 111 and the quantization control unit 112 will be described. The code amount control unit 111 distributes the allocated bit amount for each picture in the GOP based on the bit amount for a picture that has not been encoded in the GOP including the allocation target picture. This distribution is repeated in the order of the encoded pictures in the GOP, and a picture target code amount is set for each picture.

  Next, the quantization control unit 112 obtains a quantization scale reference value based on the capacity of the virtual buffer in order to match the target code amount for each picture with the actual generated code amount. More specifically, the amount of generated code for each macroblock output from the variable length coding unit 110 is obtained by feedback control. The quantization parameter used in the quantization unit 104 is determined using Expression (11) based on the quantization parameter setting instruction from the feature extraction unit 113 with respect to the reference value of the quantization scale. The above operation corresponds to steps 1 to 3 described in the background art.

  Here, the operation of the feature extraction unit 113 will be described in detail with reference to FIG. In FIG. 2, a region 200 surrounded by a thick frame is a macroblock to be encoded. Regions 201 to 204 are subblocks constituting the macroblock 200. If the macroblock is 16 × 16, the subblock is 8 × 8.

  The feature extraction unit 113 designates either the first type of quantization parameter for a region containing a lot of high-frequency components or the second type of quantization parameter for a region containing a flat part or an edge part. Generated setting instructions. Here, the first type of quantization parameter and the second type of quantization parameter have different quantization scales, and the second type of quantization parameter is set to have a smaller quantization scale. It has become.

  Regions 205 to 212 are subblocks adjacent to the encoding target macroblock 200 and have the same size as the subblocks 201 to 204. Further, FIG. 2 shows an area where the colored portion is hardly noticeable visually such as a texture. In addition, the uncolored portion indicates a region where visual deterioration such as flatness is easily noticeable. Therefore, this indicates that the encoding target macroblock 200 has an edge where visual degradation is conspicuous.

  The feature extraction unit 113 calculates the variance that is complexity for each of the sub-blocks 201 to 204 included in the encoding target macroblock 200. Here, the variance value is a value indicating the degree of variation in pixel values within a block. For example, when the pixel values in the sub-block are almost equal, the variance is small. On the other hand, when the variation in pixel values in the sub-block is large, the variance is large. Further, the lower the pixel value level, the smaller the variance of the pixel values in the block, indicating that the block has a higher tendency for a flat image.

  Also, the variance is calculated for the subblocks 205 to 212 adjacent to the encoding target macroblock 200. If the image is to be processed in raster order from the upper left, 205 to 210 are sub-blocks included in a macroblock that has already been encoded. There is no need to calculate. Since the sub-blocks 211 and 212 have not been encoded, it is necessary to perform pre-reading and calculation.

  When the minimum value is determined from the variance values of the sub-blocks 201 to 212 and the minimum value is smaller than a predetermined threshold value (second threshold value Th2 described later), the encoding target macroblock 200 is visually degraded. It is determined that there is a possibility of a noticeable flatness or an edge. If it is necessary to separate the flat and the edge, it is possible to determine that all of the dispersion values of the sub-blocks 201 to 204 are small and that the other is an edge.

  Further, the maximum value and the minimum value are determined for the calculated dispersion values of the sub-blocks 201 to 212, and if the difference between the maximum value and the minimum value is equal to or greater than a predetermined threshold (first threshold Th1 described later) The encoding target macroblock 200 can be determined as the edge portion.

  With a technique such as TM5, in which the determination is made only by the distribution of the sub-blocks 201 to 204 included in the encoding target macroblock 200, there is a possibility that the encoding target macroblock 200 is erroneously determined as a texture that is not easily noticeable visually. . On the other hand, by taking into account the dispersion of subblocks adjacent to the encoding target macroblock 200 as in the present embodiment, it is possible to accurately capture an edge without causing erroneous determination.

  Since the block determined to be an edge in this manner is a block in which visual degradation is conspicuous, an instruction to set the second type of quantization parameter is sent to the quantization control unit 112. The information transmission format may be the same as the activity. Also, even when it is determined that the encoding target macroblock is flat, a setting instruction specifying the second type of quantization parameter is sent to the quantization control unit 112.

  On the other hand, when the minimum value of the variance values of the sub-blocks 201 to 212 is equal to or greater than the second threshold value, it can be determined that the macroblock is a block containing a lot of high-frequency components that is hardly visually degraded. Therefore, an instruction to set the first type of quantization parameter having a larger quantization scale is sent to the quantization control unit 112.

  FIG. 3 is a flowchart showing an example of the quantization parameter determination process described above. Hereinafter, the flow of processing corresponding to the flowchart of FIG. 3 will be described. This processing is realized by the feature extraction unit 113 executing a corresponding processing program.

  First, in step S301, a processing target macroblock is selected. In FIG. 2, the macro block 200 is selected as the processing target macro block. Next, in step S302, a subblock adjacent to the selected macroblock is determined. In the case of FIG. 2, subblocks 205 to 212 adjacent to the macroblock 200 are determined as adjacent subblocks.

  Next, in step S303, the variance of the subblock included in the macroblock and the subblock determined in step S303 is calculated. In the example of FIG. 2, the variances of the sub-blocks 201 to 212 are calculated. In subsequent step S304, a minimum value and a maximum value are determined for the calculated distribution of sub-blocks. Also, the difference between the maximum value and the minimum value is calculated.

  In a succeeding step S305, it is determined whether or not the difference calculated in the step S304 is larger than the first threshold Th1. If the difference is larger than the first threshold Th1 (“YES” in step S305), the process proceeds to step S307. On the other hand, when the difference is equal to or smaller than the first threshold Th1 (“NO” in step S305), the process proceeds to step S306.

  In step S306, it is determined whether or not the minimum value determined in step S304 is smaller than the second threshold Th2. If the minimum value is smaller than the second threshold Th2 (“YES” in step S306), the process proceeds to step S307. On the other hand, when the minimum value is equal to or greater than the threshold Th2 (“NO” in step S306), the process proceeds to step S308.

  In step S307, an instruction to set the second type of quantization parameter is transmitted to the quantization control unit 112, and the process ends. In step S308, an instruction to set the first type of quantization parameter is transmitted to the quantization control unit 112, and the process ends.

  Note that the quantization parameter setting criteria are not limited to the above description. For example, it is determined that the variance of any of the sub-blocks 201 to 204 is smaller than the second threshold value, but the variance values of the adjacent sub-blocks 205 to 212 are all greater than the predetermined third threshold value. The case where it was calculated is also considered. Here, a sub-block having a variance smaller than the second threshold can be determined to be an isolated flat portion existing in the texture, and it can be said that visual deterioration is not noticeable. Therefore, in such a case, an instruction to set a large quantization parameter can be sent to the quantization control unit 112.

  The specific values of the above-mentioned threshold values are not particularly limited because the values themselves do not constitute the essential technical idea of the present invention. That is, each threshold value may be set to a suitable value by those skilled in the art when carrying out the invention, and the setting of the value itself is an action within the ordinary creative ability of those skilled in the art.

  As described above, according to the first embodiment, a quantization parameter for detecting a block including an edge from information of a sub-block adjacent to an encoding target macroblock and finely quantizing such a block. Can be determined. As a result, it is possible to improve the image quality of blocks that are prominent in visual deterioration.

[Second Embodiment]
Next, a second embodiment will be described with reference to FIGS. In the second embodiment, the encoding process is pipelined so that feature extraction according to the present invention can be performed efficiently. FIG. 4 is a diagram showing an example of the entire system configuration in the second embodiment of the invention. FIG. 5 is a diagram for explaining the processing operation of the system according to the second embodiment of the invention. FIG. 6 is a view similar to FIG.

FIG. 4 has a configuration in which processing is pipelined with respect to FIG. An input signal 401 is an input signal to the encoding device and is divided into macro blocks and input.
The encoding device includes an addition / subtraction unit 402, an orthogonal transform (DCT) unit 403, a quantization unit 404, an inverse quantization unit 405, an inverse orthogonal transform (inverse DCT) unit 406, a motion estimation unit 407, a motion compensation unit 408, and a frame memory. 409. Further, it includes a variable length coding unit 410, a code amount control unit 411, a quantization control unit 412, and feature extraction units 413, 414, and 415, and outputs an output signal 416.

  The operation of the encoding device corresponding to this embodiment will be described below. When the coded picture is an intra-frame coded (intra-coded) picture, the orthogonal transform unit 403 performs orthogonal transform processing on the input signal 401 divided into blocks, and the quantized unit. At 404, the DCT coefficients are quantized.

  When the encoded picture is an inter-frame encoded (inter-coded) picture, the inverse quantization unit 405 performs inverse quantization on the encoded picture, and the inverse orthogonal transform unit 406 performs inverse orthogonal transform processing. Create a local decoded image. The motion estimation unit 407 performs motion estimation with the encoding target picture, the motion compensation unit 408 performs motion compensation, and the addition / subtraction unit 402 calculates a difference value from the local decoded image. The difference value is subjected to orthogonal transform processing on the signal in the macroblock in the orthogonal transform unit 403, and the DCT coefficient is quantized in the quantization unit 404.

  Regardless of intra-frame coding or inter-frame coding, the quantized signal quantized by the quantizing unit 404 is encoded by the variable length coding unit 410 and the encoded signal is output as an output signal 416. The

  Next, operations of the code amount control unit 411, the quantization control unit 412, and the feature extraction unit 413 will be described.

  The code amount control unit 411 distributes the allocated bit amount for each picture in the GOP based on the bit amount for a picture that has not been encoded in the GOP including the allocation target picture. This distribution is repeated in the order of the encoded pictures in the GOP, and a picture target code amount is set for each picture.

  Next, the quantization control unit 412 obtains a reference value of the quantization scale based on the capacity of the virtual buffer in order to match the target code amount for each picture with the actual generated code amount. Specifically, the amount of generated code for each macroblock output from the variable length coding unit 410 is obtained by feedback control. The quantization parameter used in the quantization unit 404 is determined using Expression (11) based on the setting instruction from the feature extraction unit 413 with respect to the reference value of the quantization scale. The above operation corresponds to steps 1 to 3 described in the background art.

  Also, the coding apparatus corresponding to the present embodiment operates as a whole with a three-stage pipeline configuration. The first stage (420 in FIG. 4) is a feature extraction unit 413, and the second stage (421 in FIG. 3) is a feature extraction unit 414 and a motion estimation unit 407. The third stage (422 in FIG. 4) includes a feature extraction unit 415, a quantization control unit 412, a motion compensation unit 408, an orthogonal transformation unit 403, a quantization unit 404, a variable length coding unit 410, an inverse quantization unit 405, This is an inverse orthogonal transform unit 406.

  The reason for adopting such a pipeline configuration is as follows.

  The motion estimation unit 407 is required to have an enormous amount of calculation although it depends on the size of the search range. Considering the bus rate, etc., it is possible to use most of the time allowed for processing the macroblock. On the other hand, since the motion compensation unit 408 reads image data from the reference frame based on the motion vector determined by the motion estimation unit 407, the process cannot be started without the motion vector.

  Therefore, it is desirable that the motion vector is determined at the start of the motion compensation unit 408. In FIG. 4, the motion estimation unit 407 is configured to perform processing one step earlier than the motion compensation unit 408. The modules other than the motion compensation unit 408 belonging to the third stage are put in the same stage in consideration of the processing amount, but may be configured to be further delayed by one stage. For example, H.M. In the case of H.264 / AVC, CABAC is used for variable-length coding. However, the processing amount varies depending on the input signal, and the number of cycles may not be accurately estimated. It may be a system.

  Since the feature extraction unit 413 calculates the feature quantity for determining the quantization parameter for performing the quantization in the quantization control unit 412 prior to the quantization, the feature extraction unit 413 is also more than the quantization control unit 412. It is configured to process quickly. Further, in this configuration, the configuration is two steps earlier than the first step, which will be described in conjunction with FIG.

  In FIG. 5, (a) indicates that the processing of the (N + 2) -th macroblock (420 in FIG. 4) is performed when the quantization target macroblock is N-th. (B) shows that the (N + 1) -th macroblock processing (421 in FIG. 4) is being performed, and (c) shows that the (N) -th macroblock processing (422 in FIG. 4) is being performed. .

  In FIG. 6, reference numeral 601 denotes an encoding target macroblock, 600 is the previous macroblock (N-1 macroblock), and 602 is the next macroblock (N + 1 macroblock). Reference numerals 603 to 620 denote sub-blocks.

  When the macro block 600 on the left side of the encoding target macro block 601 is input to the encoding apparatus, the feature extraction unit 413 in the first stage 420 calculates the variance of the sub blocks 603 to 606.

  When the process proceeds by one macroblock, the encoding target macroblock 601 is input to the feature extraction unit 413 in the first stage 420, and the variance values of the subblocks 607 to 610 are calculated. Also, the second stage feature extraction unit 414 holds the variance values of the sub-blocks 603 to 606 of the macroblock 600 acquired from the feature extraction unit 413. Further, the motion estimation unit 407 performs processing for determining a motion vector for the macroblock 600.

  When the processing further advances by one macroblock, the macroblock 602 on the right side of the encoding target macroblock 601 is input to the first-stage feature extraction unit 413, and the variance values of the subblocks 611 to 614 are calculated. In addition, the second stage feature extraction unit 414 holds the variances of the sub-blocks 607 to 610 in the encoding target macroblock 601 acquired from the feature extraction unit 413. The third stage feature extraction unit 415 holds the distribution of the sub-blocks 603 to 606 of the macroblock 600.

  At this time, the dispersion values of the subblocks 607 to 610 in the encoding target macroblock 601, the subblocks 604 and 606 adjacent to the left, and the subblocks 611 and 613 adjacent to the right can be prepared.

Therefore, the process proceeds by one macroblock, and the feature extraction unit 415 calculates the minimum value and the maximum value from the variances of these subblocks and the subblocks 616 to 619 held in advance. Also, it is determined whether or not the encoding target macroblock 601 belongs to a flat part or an edge. Further, according to the determination result, a quantization parameter setting instruction is generated for the quantization control unit 412. The quantization parameter setting instruction is given to the quantization control unit 412. Note that the quantization parameter setting instruction generation processing is the same as that described with reference to the flowchart of FIG. 3 in the first embodiment. Therefore, detailed description is omitted.

  As described above, according to the present embodiment, the function of the feature extraction unit, which has been integrated into one in the first embodiment, has a three-stage pipeline configuration. Thereby, prior to the process of determining the quantization parameter by the quantization control unit 412, the distribution of sub-blocks can be calculated to generate a quantization parameter setting instruction, so that the processing efficiency can be improved. Further, the processing efficiency can be increased by performing the motion vector calculation process with a large amount of calculation prior to the process in the motion compensation unit 408.

[Other Embodiments]
The object of the present invention can also be achieved by supplying a storage medium in which a computer program code for realizing the above-described functions is recorded to a system, and the system reads and executes the computer program code. In this case, the computer program code itself read from the storage medium realizes the functions of the above-described embodiments, and the storage medium storing the computer program code constitutes the present invention. In addition, the operating system (OS) running on the computer performs part or all of the actual processing based on the code instruction of the program, and the above-described functions are realized by the processing. .

  Furthermore, you may implement | achieve with the following forms. That is, the computer program code read from the storage medium is written into a memory provided in a function expansion card inserted into the computer or a function expansion unit connected to the computer. Then, based on the instruction of the code of the computer program, the above-described functions are realized by the CPU or the like provided in the function expansion card or function expansion unit performing part or all of the actual processing.

  When the present invention is applied to the above storage medium, the computer program code corresponding to the flowchart described above is stored in the storage medium.

It is a figure which shows an example of the whole system configuration | structure in the 1st Embodiment of invention. It is a figure for demonstrating the feature extraction corresponding to the 1st Embodiment of invention. It is a flowchart which shows an example of the process of the feature extraction part 113 in the 1st Embodiment of invention. It is a figure which shows an example of the whole system configuration | structure in the 2nd Embodiment of invention. It is a figure for demonstrating the processing operation of the system in the 2nd Embodiment of invention. It is a figure for demonstrating the feature extraction corresponding to 2nd Embodiment.

Explanation of symbols

101 ... Input unit (input signal)
102 ... Addition / subtraction 103 ... Orthogonal transformation unit 104 ... Quantization unit 105 ... Inverse quantization unit 106 ... Inverse orthogonal transformation unit 107 ... Motion estimation unit 108 ... Motion compensation unit 109 ... Frame memory 110 ... Variable length encoding unit 111 ... Code amount control unit 112 ... Quantization control unit 113 ... Feature extraction unit 116 ... Output unit (output signal, stream)

Claims (7)

An encoding device that performs encoding by dividing an image into a plurality of macroblocks,
Generating means for generating a setting instruction for specifying a quantization parameter used for quantization for each of the plurality of macroblocks;
Determining means for determining a quantization parameter of the macroblock based on the setting instruction generated by the generating means;
Transform means for performing orthogonal transformation of the macroblock and quantization using the determined quantization parameter;
Encoding means for variable-length encoding the conversion result by the conversion means;
With
The generating means calculates a variance for subblocks constituting the macroblock, and based on the variance of subblocks constituting the macroblock and the variance of subblocks adjacent to the macroblock, the quantization parameter An encoding device that generates a setting instruction. In the setting instruction,
A first type of quantization parameter for quantization of a macroblock containing a lot of high frequency components;
2. The encoding apparatus according to claim 1, wherein one of a second type quantization parameter having a quantization scale smaller than the first type quantization parameter is designated.   The generating means determines a minimum value and a maximum value among a variance of subblocks constituting the macroblock and a variance of subblocks adjacent to the macroblock, and a difference between the maximum value and the minimum value is determined. The encoding apparatus according to claim 2, wherein a setting instruction that specifies the second type of quantization parameter is generated when greater than a first threshold value. The generating means determines a minimum value among a variance of subblocks constituting a macroblock and a variance of subblocks adjacent to the macroblock;
When the minimum value is smaller than a second threshold, a setting instruction for specifying the second type of quantization parameter is generated;
4. The encoding apparatus according to claim 2, wherein when the minimum value is not smaller than the second threshold, a setting instruction for specifying the first type of quantization parameter is generated. 5.   The generating means, when the variance of any of the sub-blocks constituting the macro block is smaller than the second threshold and the variance of the sub-blocks adjacent to the macro block is larger than a third threshold, The encoding apparatus according to claim 4, wherein a setting instruction for specifying a second type of quantization parameter is generated. A method for controlling an encoding device that performs encoding by dividing an image into a plurality of macroblocks,
For each of the plurality of macroblocks, a generation step for generating a setting instruction that specifies a quantization parameter used for quantization;
A determination step of determining a quantization parameter of the macroblock based on the setting instruction generated in the generation step;
A transform step for performing orthogonal transformation of the macroblock and quantization using the determined quantization parameter;
An encoding step for variable-length encoding the conversion result in the conversion step;
With
In the generating step, a variance is calculated for subblocks constituting the macroblock, and the quantization parameter is calculated based on a variance of subblocks constituting the macroblock and a variance of subblocks adjacent to the macroblock. A control method for an encoding device, characterized in that a setting instruction is generated.   A computer program for causing a computer to function as the encoding device according to any one of claims 1 to 5.