JP2009111712A - Video encoding device and method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a video encoding device and method for improving the quality of video by reducing a decrease in encoding efficiency while performing effective slice division for decentralization of encoding and decoding processes. <P>SOLUTION: The video encoding device has a block encoding decision unit which decides kinds of encoding of blocks to be encoded included in input video data, a slice division unit which divides a picture included in the input video data into slices comprising one or more blocks to be encoded, and a block encoding unit which encodes the input video data by the blocks to be encoded, and the slice division unit changes the number of slices constituting the picture in accordance with the distribution of encoding kinds of the blocks to be encoded. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to a video encoding apparatus and method for compressing and encoding video data.

  In recent years, with the advancement of digital signal processing technology, it has become possible to efficiently encode a large amount of digital information such as moving images, still images, and voices, and to perform recording on a small recording medium or transmission via a communication medium. . By applying such technology, video encoding devices capable of converting TV broadcast and video camera images into streams have been developed.

  MPEG (Moving Picture Experts Group) -2 and MPEG-4 are known as high-efficiency compression encoding systems for storing or transmitting video information. Furthermore, MPEG-4 Part-10: AVC is standardized. MPEG4 Part-10: AVC (ISO / IEC 14496-10) is an H.264 standard. It is also called H.264. These compression methods are compression methods using correlation in moving images, and realize a high compression rate with little deterioration.

  H. In the H.264 system, each image (frame or field) included in input video information is handled as a processing unit called a picture, and this picture can be divided into one or more slices. Further, the slice is divided into units of one or more macroblocks for encoding processing. Although it is possible to handle one picture as one slice, that is, to encode a picture without dividing the picture into slices, for the purpose of distributing the encoding process, one picture is divided into two or more pictures. There are many cases where encoding is performed by dividing into slices. For example, when decoding encoded data using a multi-core type CPU that is the mainstream in recent years, efficient decoding processing is possible by dividing a picture into multiple slices and assigning each slice to each core It is.

In addition, as a technique for dividing a picture into slices, a proposal has been made to classify image blocks by determining an image pattern of image blocks in a picture and configure a plurality of slices based on the classification result (for example, , See Patent Document 1).
JP 2006-114979 A

  FIG. 6 is a diagram illustrating an example of slice division. 900 is a picture, and 901 to 904 are slices. As shown in FIG. 6, in general slice division, a picture is divided into strips at substantially equal intervals. Each slice is composed of one or a plurality of macroblocks in the horizontal direction or the vertical direction. 905 and 906 in FIG. 6 are examples of macroblocks included in the picture 900.

  By the way, for example In the H.264 system, there is a problem in that the correlation between macroblocks straddling slice boundaries cannot be utilized for encoding. That is, when the macroblock 905 of FIG. 6 is encoded, the correlation with the macroblock 906 that touches the upper side across the slice boundary cannot be used. H. In the H.264 coding, effective coding is possible by intra prediction from adjacent macroblocks in intra coding. However, when slice division is performed as shown in FIG. May not be sufficiently utilized, and encoding efficiency may be reduced. Further, when the slice is divided by the method of Patent Document 1 described above, there is a possibility that the vertical and horizontal correlations cannot be fully utilized. When the correlation between the macroblocks is lowered in this way, a large amount of code is required for the macroblock in order to maintain the image quality.

  Therefore, in the conventional method, although the effect of distribution of encoding and decoding processing can be expected by slice division, the encoding processing efficiency may be reduced and the video quality may be reduced.

  In consideration of such a problem, the present invention performs video slice that reduces the decrease in coding efficiency and improves the video quality while performing effective slice division for distribution of coding and decoding processing. An object is to provide an apparatus and method.

  In order to achieve the above object, a video encoding device of the present invention is a video encoding device that encodes video data, and determines an encoding type of an encoding target block included in input video data. Determination means for dividing, a dividing means for dividing a picture included in the input video data into slices composed of one or more blocks to be encoded, and a slice divided by the slice dividing means are encoded. And encoding means for encoding the input video data for each of the encoding target blocks, and the slice dividing means is an encoding type of the encoding target block determined by the determination means The number of slices constituting the picture is changed in accordance with the distribution of.

  The video encoding method of the present invention is a video encoding method for encoding video data, and a determination step of determining an encoding type of an encoding target block included in input video data; A slice dividing step of dividing a picture included in the input video data into slices composed of one or more encoding target blocks, and the input video data as the encoding unit. An encoding step for encoding each block to be encoded, and in the slice division step, the picture is configured according to a distribution of encoding types of the block to be encoded determined in the determination step It is characterized by changing the number of slices.

  According to the present invention, it is possible to perform slice division effective for distribution of encoding and decoding processes, reduce a decrease in encoding efficiency, and improve video quality.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.

  FIG. 1 is a block diagram of a video encoding apparatus according to the present embodiment. In FIG. 1, 101 is a video data input unit, 102 is a block coding determination unit, 103 is a slice division unit, 104 is a block coding unit, and 105 is a stream output unit. The stream output unit 105 outputs encoded data (stream).

  Picture data input from the video data input unit 101 is supplied to the block encoding determination unit 102 and the slice division unit 103. The block coding determination unit 102 determines the coding type for each block constituting the input picture data. The block here is a unit of encoding processing such as a macroblock including a plurality of pieces of pixel information. In addition, the determination of the encoding type is, for example, whether the block to be encoded is a block to be inter-encoded using the correlation with other pictures or an intra-frame using the correlation between adjacent blocks in the picture. It is to determine whether the block is to be encoded. Further, the block coding determination unit 102 also determines a pixel sequence to be correlated of the encoding target block, a correlation calculation method, and the like. The determination result of the block encoding determination unit 102 is supplied to the slice dividing unit 103 and the block encoding unit 104 as information for encoding.

  Whether the slice division unit 103 should divide the picture into slices based on the distribution of the coding type of each block in the picture based on the determination result of the block coding determination unit 102 with respect to the picture data input from the video data input unit 101 Decide whether or not. For example, if it is determined that blocks exceeding a predetermined ratio (for example, 40%) among the encoding target blocks in the picture should be intra-encoded, the picture is not divided into slices in consideration of the cross-slice correlation. Set to be a single slice. On the other hand, for example, if the number of blocks determined to be intra-encoded is less than a predetermined ratio (for example, 40% or less), the slice division is set in consideration of the distribution of encoding and decoding processes. Alternatively, the number of slice divisions may be determined according to the ratio of blocks determined to be intra-coded to the whole. Thus, the number of slice divisions changes according to the ratio of the encoding target blocks that are intra-encoded. A slice includes one or more encoding target blocks in a picture, and the entire picture can be a single slice.

  In this manner, the slice division unit 103 performs slice division setting on the picture data supplied from the video data input unit 101 based on the determination result of the block encoding determination unit 102, and the picture data is converted into the block encoding unit. 104 is supplied. Further, the slice dividing unit 103 supplies information such as the position of the block to be divided into slices to the block encoding unit 104 together with the picture data as information for encoding.

  The block encoding unit 104 encodes the picture data supplied from the slice dividing unit 103 for each block according to the encoding type information (block encoding determination information) of each block supplied from the block encoding determination unit 102. Turn into. When the input picture data is divided into slices, the block encoding unit 104 performs encoding in units of each slice according to the set slice division information (slice division setting information). Further, the block encoding unit 104 outputs encoded data (stream) obtained as a result of the encoding process from the stream output unit 105.

  Here, details of the block encoding unit 104 will be described with reference to FIG. FIG. 5 is a block diagram showing the block encoding unit 104 in detail.

  In FIG. 5, the block encoding unit 104 performs an encoding process for each input encoding target block. The block encoding unit 104 receives the data of the encoding target block, and also inputs the block encoding determination information supplied from the block encoding determining unit 102 and the slice division setting information supplied from the slice dividing unit 103 Is done.

  When performing the predictive coding, the subtractor 14 calculates a difference between the input block data (image data) to be encoded and the predicted value from the switch 38, and uses the difference data as an integer precision discrete cosine. It outputs to the integer conversion part 16 which performs conversion (DCT). The subtracter 14 outputs the input image data to the integer conversion unit 16 as it is when the predictive encoding is not performed.

  The integer transform unit 16 performs discrete cosine transform on the image data from the subtractor 14 with integer precision, for example, in units of 4 × 4 pixel blocks, and outputs the resulting DCT transform coefficients to the quantization unit 18.

  The quantization unit 18 performs a quantization process on the DCT transform coefficient from the integer transform unit 16 and outputs the obtained quantized transform coefficient to the entropy coding unit 40 and the inverse quantization unit 20.

  The entropy encoding unit 40 entropy encodes the quantized transform coefficient from the quantization unit 18. For example, H.C. In the case of the H.264 encoding method, CABAC (Context-Based Adaptive Binary Arithmetic Coding) can be used as entropy encoding. Alternatively, CAVLC (Context-Based Adaptive Variable Length) can also be used. The entropy encoding unit 40 performs an encoding process using a slice as a basic unit based on the input slice division setting information, for example, by the CABAC method. The entropy encoding unit 40 also multiplexes the motion vector information from the motion estimation unit 36 with the encoded data, and outputs the encoded data (stream). It should be noted that information such as the position of the slice-divided block and the coding type (I slice, P slice, or B slice) of the slice is also multiplexed in the output stream and can be used in the decoding process.

  The inverse quantization unit 20 returns the quantization variable coefficient from the quantization unit 18 to the DCT transform coefficient and sends it to the inverse integer transform unit 22. The inverse integer transform unit 22 performs an inverse integer transform process on the DCT transform coefficient from the inverse quantization unit 20 to restore data corresponding to the output data of the subtractor 14.

  The adder 24 outputs the output data of the inverse integer transform unit 22 as it is when the predictive encoding is not used, and the output data of the inverse integer transform unit 22 when the predictive encoding is used. The predicted value from the switch 38 is added to and output. The output data of the adder 24 is so-called local decoded data. The output of the adder 24 is stored in the frame memory 26 for intra coding by intra-picture prediction. Further, it is stored in the frame memory 32 via the deblocking filter 30 for inter coding by inter-picture prediction. The deblocking filter 30 is used to reduce block noise in the locally decoded data.

  The motion estimation unit 36 estimates the motion of the object in the image by comparing the input encoding target block data with the locally decoded data stored in the frame memory 32 in relation to the inter prediction described later. The locally decoded data stored in the frame memory 32 constitutes a picture located before or after the display timing of the current picture, and is used as a reference image for prediction between pictures. The motion vector detected by the motion estimation unit 36 is supplied to the inter prediction unit 34 and the entropy encoding unit 40.

  In addition, the motion estimation unit 36 operates to prohibit detection of a motion vector that exceeds the boundary between slices based on the input slice division setting information. Furthermore, the motion estimation unit 36 can also select a reference image and a reference pixel sequence suitable for the encoding target block using the input block coding determination information.

  The intra prediction unit 28 refers to the locally decoded data stored in the frame memory 26, performs a prediction process, calculates a prediction value from the same picture, and outputs it. When the picture is divided into slices, the intra prediction unit 28 performs prediction processing using only information in the picture and the same slice based on the input slice division information, and calculates a prediction value. The intra prediction here is to generate a prediction value of a prediction target block using information on adjacent pixels included in an encoded pixel block.

  The inter prediction unit 34 determines image data to be used for prediction from the reference image stored in the frame memory 32 according to the motion vector estimated by the motion estimation unit 36, and predicts inter-picture prediction values (so-called motion compensation prediction). Value) is calculated and output.

  The switch 38 uses the prediction value of the intra prediction unit 28 or the inter prediction unit 34 depending on whether the block to be encoded is encoded by intra prediction (intra encoding) or encoded by inter prediction (inter encoding). Select the predicted value of. The selection result of the switch 38 is supplied to the subtractor 14 as a predicted value of predictive coding. The switch 38 selectively switches between intra coding and inter coding based on the input block coding determination information. However, depending on the position of the slice division, intra coding or inter coding specified by the block coding determination information may not be suitable. In such a case, the output is switched to an appropriate one.

  Thus, according to the present embodiment, the presence or absence of slice division is adaptively switched according to the distribution of the coding types of blocks (for example, macroblocks) constituting a picture. By switching from picture to picture, such as no division when coding efficiency is likely to be sacrificed by slice division, and there is division when the effect on coding efficiency is small, the coding efficiency of the entire stream is improved. Processing dispersion by slice division can be realized while maintaining.

  Another embodiment of the above-described slice division determination will be described. FIG. 2 is a diagram illustrating a configuration example of slices and blocks in a picture. In FIG. 2, 200 is a picture, and 201 to 204 are slices assumed at the time of division. Reference numerals 205 to 207 denote blocks (for example, macroblocks) that are in contact with the lower side of each assumed slice division line.

  In the present embodiment, the presence / absence of slice division is determined according to the proportion of blocks determined to be intra-coded in the blocks 205 to 207 to be coded. Further, instead of determining the division in all blocks 205 to 207, the distribution of 205, 206, and 207 may be individually determined according to the division position, and the presence / absence of division may be determined for each division position.

  Further, more effective control can be achieved by determining whether or not to divide from the ratio of intra coding using upper and lower references, instead of simply obtaining the ratio of intra coding.

  FIG. FIG. 2 is a diagram illustrating a sequence of pictures in the H.264 encoding scheme. 301 is an I picture, 302 is a P picture, and 303 is a B picture. In the I picture, all blocks are intra-coded. For P and B pictures, inter coding using the correlation with the preceding and following pictures may be used, or intra coding using the correlation within the picture may be used.

  In the I picture 301 in FIG. 3, the ratio of intra-coded blocks is 100%. For this reason, obviously, an I picture is highly likely to be able to use correlation across slices by intra coding. By using such properties, in the configuration of the first embodiment described above, it is preferable to set so that slice division is not always performed (a single slice) for an I picture. By configuring as described above, more efficient control becomes possible. Such a configuration is also within the scope of the present invention.

  In the first embodiment described above, the description has been given focusing on the upper and lower correlations of intra coding. However, in other coding types, the correlation between the upper and lower blocks (for example, macroblocks) affects the coding efficiency. If given, it can be used to determine slice division. For example, H.M. In H.264, the correlation between the upper and lower blocks is also used in the coding of the motion vector used for the inter prediction in the inter coding.

  FIG. 4 is a diagram illustrating a configuration example of slices and blocks in a picture. In FIG. 4, 400 is a picture, and 401 to 404 are slices assumed at the time of division. Reference numerals 405 to 408 denote encoding target blocks (for example, macro blocks), respectively.

  In the example of FIG. 4, the motion vectors of the blocks 405 and 406 that are in contact with each other with the slice division position (slice boundary) interposed therebetween are approximate, and the correlation is high. On the other hand, the blocks 407 and 408 have a low correlation of motion vectors even under the same arrangement condition. When the number of combinations with high correlation exceeds a considerable number as in the former case, the slice division at this position should have a significant impact on the coding efficiency.

  Therefore, the slice division unit 103 described above determines whether or not to perform slice division (presence / absence of block division) in consideration of the correlation of motion vectors between blocks sandwiching a position (boundary part) that assumes slice division. Good. That is, at a position (boundary portion) where slice division is assumed, control is performed such that a block having a high correlation exceeds 40%, for example, and is not divided when it is 40% or less. By configuring as described above, more effective control becomes possible.

  In each of the above embodiments, the video encoding device that encodes the video data supplied from the video data input unit 101 has been described as an example based on the block diagram shown in FIG. Instead of the unit 101, a configuration having an imaging unit may be used. The imaging unit outputs video data obtained by photographing the subject, and the video data is encoded by the video encoding device described above. That is, it is also within the scope of the present invention to mount the video encoding apparatus of the present embodiment in an imaging apparatus so as to function as a part of the imaging apparatus.

It is a block diagram of the video coding apparatus which concerns on embodiment of this invention. It is a figure which shows the structural example of the slice and block in a picture. It is a figure which shows the arrangement | sequence of a picture. It is a figure which shows the structural example of the slice and block in a picture. It is a block diagram showing the block encoding part in detail. It is a figure which shows an example of a slice division | segmentation.

Explanation of symbols

DESCRIPTION OF SYMBOLS 101 Video data input part 102 Block encoding determination part 103 Slice division part 104 Block encoding part

Claims (10)

A video encoding device for encoding video data,
Determining means for determining an encoding type of an encoding target block included in the input video data;
Division means for dividing a picture included in the input video data into slices composed of one or more encoding target blocks;
An encoding unit that encodes the input video data for each of the encoding target blocks, with the slice divided by the slice dividing unit as an encoding unit;
The video coding apparatus characterized in that the slice dividing means changes the number of slices constituting the picture in accordance with the distribution of the coding type of the coding target block determined by the determining means.   The video encoding apparatus according to claim 1, wherein the encoding type indicates that the encoding target block is intra-encoded.   The video encoding apparatus according to claim 1, wherein the encoding type indicates that the encoding target block is inter-encoded.   The slice dividing unit is characterized in that, in the picture, when the coding target block determined to be intra-coded by the determining unit exceeds a predetermined ratio, the picture is made a single slice. Item 2. The video encoding device according to Item 1.   The slice dividing unit divides the picture into a plurality of slices when a coding target block determined to be intra-coded by the determining unit is equal to or less than a predetermined ratio in the picture. Item 5. The video encoding device according to Item 1 or 4.   The slice dividing unit determines whether or not to perform block division according to a ratio at which a coding target block adjacent to a lower side of a division position assuming slice division is determined to be intra-coded by the determination unit. The video encoding device according to claim 1.   2. The video encoding apparatus according to claim 1, wherein, when the picture is an I picture, the slice dividing unit sets the picture as a single slice.   2. The video encoding apparatus according to claim 1, wherein the slice dividing unit determines presence / absence of block division according to a correlation of a plurality of encoding target blocks that are in contact with each other above and below a division position assuming slice division. .   The encoding unit selectively performs intra encoding and inter encoding of the encoding target block using determination information supplied from the determination unit and slice division setting information supplied from the slice division unit. The video encoding device according to claim 1, wherein the video encoding device is a video encoding device. A video encoding method for encoding video data,
A determination step of determining an encoding type of an encoding target block included in the input video data;
A slice dividing step of dividing a picture included in the input video data into slices composed of one or more encoding target blocks;
An encoding step of encoding the input video data for each of the encoding target blocks using the slice as an encoding unit;
The video coding method characterized in that, in the slice division step, the number of slices constituting the picture is changed in accordance with the distribution of the coding type of the coding target block determined in the determination step.

Images