JP4453398B2 - Encoding apparatus, program, and encoding processing method - Google Patents

Description

  The present invention relates to an encoding device, a program, and an encoding processing method that perform encoding processing such as arithmetic encoding on data such as image data and audio data.

  In recent years, image data has been handled as digital data. At that time, MPEG (compressed by orthogonal transform such as discrete cosine transform and motion compensation is used for the purpose of efficient transmission and storage of information, and using redundancy unique to image information. A device conforming to a system such as Moving Picture Experts Group) is becoming popular in both information distribution at broadcasting stations and information reception in general households.

Following this MPEG system, an encoding system called JVT (Joint Video Team) that realizes a higher compression rate has been proposed (see, for example, Non-Patent Document 1).
In the JVT system, two types of encoding processing, CAVLC (Context Based Adaptive Variable Length Coding) and CABAC (Context Adaptive Binary Arithmetic Coding), are defined as encoding processing of a syntax element (SE).
CABAC is an encoding process in which syntax elements (SE) are divided into contexts, and arithmetic encoding processes are performed on the respective contexts.
Thomas Wiegand, Gary J. Sullivan, Gisle Bjontegaard, and Ajay Luthra, Overview of the Video Coding Standard (Overview of the H.264 / AVC Video Coding Standard), "Processing Circuits and Systems for IEEE Video Technology" (IEEE TRANSACTIONS CIRCUITS AND SYSTEMS FOR VIDEO TECHNOLOGY), (USA), July 2003,

In the arithmetic coding process by CABAC described above, a function called CabacZeroWord for limiting the compression rate in consideration of the load on the decoder is known.
Specifically, when encoded data having a high compression rate is generated as a result of the encoding processing by CABAC, if the decoder performs decoding processing on the encoded data having a high compression rate, the encoded data is stored within a predetermined time. There are cases where it cannot be fully processed.
Therefore, at the time of encoding processing, for example, additional data is added to the end of each of a plurality of slices (Slice) in one image. Specifically, for example, the additional data adds a stuffing byte (Stuffing Byte) with 0x000003 as one unit, thereby reducing the load on the decoder. Additional data corresponds to CabacZeroWord.

However, in the encoding process described above, the data length of the data added to the end of one slice by CabacZeroWord is uncertain until the encoding process for one slice of data is completed. There are cases where the quantization rate at the time of quantization processing cannot be controlled.
As a result of the encoding process, encoded data with a high compression rate is generated, and if a large amount of CabacZeroWord occurs suddenly, the buffer may overflow.
For this reason, an encoding apparatus that can predict additional data in a processing unit shorter than one slice is desired.

  The present invention has been made in view of such circumstances, and an object of the present invention is to provide an encoding device, a program, and an encoding method capable of predicting additional data in a unit smaller than a predetermined unit of additional data in image data in encoding processing, and An object of the present invention is to provide an encoding processing method.

  Another object of the present invention is to provide an encoding device, a program, and an encoding processing method capable of predicting additional data in predetermined units of image data and performing quantization processing with high accuracy based on the predicted data. There is to do.

In order to achieve the above object, an encoding apparatus according to a first aspect of the present invention includes an arithmetic encoding unit that performs arithmetic encoding processing on image data in units of macroblocks to generate encoded data, and the arithmetic encoding means the generated encoded data, said arithmetic coding means is generated I by the arithmetic coding process, index data indicating the degree of compression of the image data in the slice or picture unit including a plurality of macroblocks adding means for adding additional data corresponding to, in the macro-block unit, the data length of the target data for arithmetic coding by the arithmetic coding means and the coded data said arithmetic coding means to generate based on the data length, and predicting means for predicting the additional data to which the additional means for adding, at the macro-block basis, the additional data the prediction means predicts And a processing means for performing predetermined processing corresponding.

Furthermore, in order to achieve the above object, the encoding apparatus according to the second aspect of the present invention refers to the image data to be subjected to motion compensation and the motion compensation in units of a plurality of macro blocks having different sizes. Motion vector generating means for generating a motion vector based on the difference between the reference image data, and the difference between the motion compensation target image data and the predicted image data are subjected to orthogonal transform processing and quantization processing in order. A first processing unit; a second processing unit configured to sequentially perform an inverse quantization process and an inverse orthogonal transform process on the data generated by the first processing unit; and generate the reference data; the motion vector and the reference A third processing unit that generates the predicted image data based on the image data, a quantization process based on the set data indicating the degree of quantization, and a process of the encoding unit; Quantization means for generating target data; data generated by the quantization means; and arithmetic coding means for generating encoded data by performing arithmetic coding processing of the motion vector in units of the macroblocks; Index data indicating the degree of compression of the image data in units of slices or pictures including a plurality of macroblocks generated by the arithmetic encoding process of the arithmetic encoding unit in the encoded data generated by the arithmetic encoding unit An additional means for adding additional data in accordance with the data length of the data to be subjected to arithmetic coding processing by the arithmetic coding means, and the encoded data generated by the arithmetic coding means in the macroblock unit. Prediction means for predicting additional data corresponding to the index data added by the adding means based on the data length; In click units, and a the prediction means on the basis of the additional data predicted, quantized data setting means for setting data indicating the degree of quantization of the quantization means.

Furthermore, in order to achieve the above object, a program according to a third aspect of the present invention is a program that is executed by an information processing apparatus that includes an encoding unit, an adding unit, a prediction unit, and a processing unit. The encoding means performs arithmetic encoding processing on image data in units of macroblocks to generate encoded data, and the adding means adds the encoded data generated by the arithmetic encoding means to the encoded data, A second procedure of adding additional data corresponding to index data indicating the degree of compression of the image data in units of slices or pictures including a plurality of macroblocks generated by the arithmetic encoding process of the arithmetic encoding means And the prediction means generates, in the macroblock unit, the data length of the data subject to the arithmetic coding process in the first procedure, and the first procedure. A third procedure for predicting additional data to be added in the second procedure based on a data length of encoded data, and the addition predicted by the processing unit in the third step in units of the macroblocks; And causing the information processing apparatus to execute a fourth procedure for performing predetermined processing according to data.

Furthermore, in order to achieve the above object, an encoding processing method according to a fourth aspect of the present invention is an encoding performed by an information processing apparatus having an encoding unit, an adding unit, a prediction unit, and a processing unit. A processing method, wherein the encoding means performs arithmetic encoding processing of image data in units of macroblocks to generate encoded data, and the adding means is generated by the arithmetic encoding means. Additional data according to index data indicating the degree of compression of the image data in units of slices or pictures including a plurality of macroblocks generated by the arithmetic encoding process of the arithmetic encoding unit is added to the encoded data. And the prediction means includes, in units of the macroblock, a data length of data to be subjected to arithmetic coding processing in the first step, and the first A third step of predicting additional data to be added in the second step based on the data length of the encoded data generated in the step; and the processing means in the third step in units of the macroblocks. And a fourth step of performing a predetermined process according to the predicted additional data.

  According to the present invention, it is possible to provide an encoding device, a program, and an encoding processing method capable of predicting additional data in a unit smaller than a predetermined unit of additional data in image data.

  In addition, it is possible to provide an encoding device, a program, and an encoding processing method that can perform quantization processing with high accuracy based on predicted data.

  Hereinafter, a JVT encoding apparatus employing an encoding process according to an embodiment of the present invention will be described.

FIG. 1 is a conceptual diagram of a communication system 1 that employs an encoding apparatus according to an embodiment of the present invention.
As shown in FIG. 1, a communication system 1 according to the present embodiment includes an encoding device 2 provided on the transmission side and a decoding device 3 provided on the reception side.
In the communication system 1, the encoding device 2 on the transmission side generates frame image data (bit stream) compressed by orthogonal transformation such as discrete cosine transformation and Karhunen-Labe transformation and motion compensation, and modulates the frame image data. Later, it is transmitted via a transmission medium such as a satellite broadcast wave, a cable TV network, a telephone line network, or a mobile phone line network.
On the receiving side, after demodulating the received image signal, frame image data expanded by inverse transformation of orthogonal transformation and motion compensation at the time of modulation is generated and used.
The transmission medium may be a recording medium such as an optical disk, a magnetic disk, and a semiconductor memory.
The decoding device 3 shown in FIG. 1 performs decoding corresponding to the encoding of the encoding device 2 and has the same configuration as the conventional one.

Hereinafter, the encoding device 2 shown in FIG. 1 will be described.
FIG. 2 is an overall configuration diagram of the encoding device 2 shown in FIG.
As shown in FIG. 2, the encoding device (information processing device) 2 includes, for example, an A / D conversion circuit 22, a screen rearrangement circuit 23, an arithmetic circuit 24, an orthogonal transformation circuit 25, a quantization circuit 26, and an encoding circuit. (Also referred to as an encoding device) 27, buffer 28, inverse quantization circuit 29, inverse orthogonal transform circuit 30, rate control circuit 32, frame memory 34, deblock filter 38, and motion prediction / compensation circuit 39 as main components Have.
The quantization circuit 26 and the rate control circuit 32 correspond to processing means according to the present invention.

Hereinafter, components of the encoding device 2 will be described.
The A / D conversion circuit 22 converts the original image signal (image data) S10 composed of the input analog luminance signal Y and color difference signals Cb and Cr into digital frame data S22, and this is a screen rearrangement circuit. To 23.
The screen rearrangement circuit 23 rearranges the frame data S22 input from the A / D conversion circuit 22 in the encoding order according to the GOP (Group Of Pictures) structure composed of the picture types I, P, and B. The frame data S23 is output to the arithmetic circuit 24, the arithmetic circuit 24, the rate control circuit 32, and the motion prediction / compensation circuit 39.

The arithmetic circuit 24 calculates the difference between the motion compensation block MCB (also referred to as a macro block) to be processed in the frame data S23 and the motion compensation block MCB of the predicted image data PI input from the motion prediction / compensation circuit 39 corresponding thereto. Is generated and output to the orthogonal transformation circuit 25.
The orthogonal transformation circuit 25 performs orthogonal transformation such as discrete cosine transformation and Karhunen-Labe transformation on the image data S24 to generate image data (for example, DCT coefficient) S25, and outputs this to the quantization circuit 26.
The quantization circuit 26 performs a quantization process on the image data S25 based on the data indicating the degree of quantization set by the rate control circuit 32, and generates data to be processed by the encoding circuit 27.
Specifically, the quantization circuit 26 quantizes the image data S25 with the quantization scale input from the rate control circuit 32 to generate image data S26, and outputs this to the encoding circuit 27 and the inverse quantization circuit 29. To do.

The encoding circuit 27 performs arithmetic encoding processing based on the image data S26, a context element (SE: Syntax Element) indicating a motion vector, and stores the image data as a processing result in the buffer 28.
At this time, for example, when inter prediction encoding is performed, the encoding circuit 27 encodes the motion vector MV input from the motion prediction / compensation circuit 39 and stores it in the header data.

The image data stored in the buffer 28 is transmitted after being modulated or the like.
The inverse quantization circuit 29 generates a signal obtained by inversely quantizing the image data S26 of the motion compensation block MCB of the reference image data referenced from the other motion compensation block MCB, and outputs the signal to the inverse orthogonal transform circuit 30.
The inverse orthogonal transform circuit 30 writes the image data S30 generated by performing the inverse transform of the orthogonal transform in the orthogonal transform circuit 25 to the data input from the inverse quantization circuit 29, and writes the image data S30 to the frame memory 34 via the deblock filter 38.
The deblock filter 38 outputs the image data from which the block distortion of the image data S30 is removed to the frame memory 34.

For example, the rate control circuit 32 finely quantizes a portion with high complexity in the image and coarsely quantizes a portion with low complexity in the image based on the frame data S23 input from the screen rearrangement circuit 23. The quantization parameter Qp is generated.
Then, the rate control circuit 32 generates a quantization scale based on the generated quantization parameter Qp and the image data read from the buffer 28, and outputs this to the quantization circuit 26.
Further, the rate control circuit 32 outputs the generated quantization parameter Qp to the motion prediction / compensation circuit 39.

  The motion prediction / compensation circuit 39 searches the search range SR in the reference image data based on the frame data S23 to be encoded and the reference image data REF from the frame memory 34, and performs motion prediction / compensation processing. A motion vector MV is calculated using the motion compensation block MCB as a unit.

At this time, the motion prediction / compensation circuit 39 encodes an I picture, a P picture, and a B picture. Here, the I picture indicates image data that is encoded only from the information of the I picture and does not perform inter-frame prediction (inter prediction encoding).
The P picture indicates image data that is encoded by performing prediction based on the previous (previous) I picture or P picture.
The B picture indicates image data encoded by bi-directional prediction based on an I picture and a P picture whose display order is the previous (past) and subsequent (future).

As the block size of the motion prediction block (macroblock MB), sizes of 16 × 16, 8 × 16, 16 × 8, and 8 × 8 pixels are defined, and 8 × 8, 4 × 8, 8 × 4, and so on. And a size of 4 × 4 pixels is defined.
The motion prediction / compensation circuit 39 specifies the position of the motion compensation block MCB having the smallest difference from the motion compensation block MCB to be processed in the frame data S23 within the search range SR of the reference image data REF, and the positional relationship between these positions. The motion vector MV corresponding to is calculated.
Further, the motion prediction / compensation circuit 39 generates predicted image data PI for the motion compensation block MCB based on the selected motion vector MV.
The motion prediction / compensation circuit 39 outputs the selected motion vector MV to the encoding circuit 27, and outputs the generated predicted image data PI to the arithmetic circuit 24.

Hereinafter, the encoding process of the encoding circuit 27 will be described in detail.
In the JVT encoding process, two types of CAVLC and CABAC are defined as encoding methods of a context element (SE). CABAC generally compresses SE more efficiently than CAVLC. The encoding circuit 27 according to the present embodiment performs encoding processing by CABAC.

In CABAC defined by JVT, SE is converted into a symbol string by VLC or fixed length code. Each symbol of the converted symbol string is called bin (binary symbol).
In the encoding process, bin is input and arithmetic encoding (AC) is performed. Based on the processing result, output bits are determined and output to the bit stream.

FIG. 3 is a diagram for explaining the arithmetic encoding process.
Arithmetic encoding processing, for example, projects a data sequence to be encoded, for example, a symbol or binary (0 and 1) series to an interval [0, 1] according to the appearance probability, and creates a probability space on the number line. Encode the binary representation with the number in the interval.
For example, in the arithmetic encoding process, as shown in FIG. 3, a section of 0.00... Or more and less than 1.00... Is divided based on the appearance probability of individual data in the encoding target data. A recursive process of selecting a section divided based on individual data is performed, and data indicating a section corresponding to the data to be encoded is used as a code.

Hereinafter, in order to simplify the description, the data to be encoded is binary.
Specifically, as shown in FIG. 3A, the arithmetic coding process is performed by using MPS (Most Probable Symbol) for data with a high appearance probability and LPS (Least for data with a low appearance probability). Probable Symbol), and a predetermined section A is divided according to the appearance probabilities pMPS and pLPS. At this time, the size of the current section is set to Range, and the boundary value of the current section, for example, the lower limit value is set to Low. The appearance probability pMPS is 1-pLPS.

In the initial state, as shown in FIG. 3A, the values of the range Range and low Low are A, 0.00.
When the first data of the data string to be encoded is MPS, as shown in FIG. 3A, the section A is divided according to the appearance probabilities pMPS and pLPS, and among the divided sections, FIG. As shown in (B), the section corresponding to MPS is selected. The values of Range Range and Low Low at this time are p0 (= pMPS), 0.00.

  Next, when the second data in the data string to be encoded is MPS, the section is divided according to the appearance probabilities pMPS and pLPS as shown in FIG. Of these, the section corresponding to the MPS is selected as shown in FIG. The values of the range Range and low Low at this time are p00 (= p0 × pMPS), 0.00.

Next, when the third data in the data string to be encoded is LPS, the section is divided according to the appearance probabilities pMPS and pLPS as shown in FIG. Of these, the section corresponding to LPS is selected as shown in FIG. The values of Range Range and Low Low at this time are p001 (= p00 × pLPS) and p001 (Low + pMPS).
That is, in the case of LPS in the above-described encoding process, pMPS is added to the value of the lower limit value Low, and the lower limit value Low is updated.

  Further, as the bit length of the data string to be encoded increases, the range Range decreases and the number of bits of data indicating the range Range also increases. For this reason, the capacity of the memory is reduced by outputting determined bits as a result of the encoding process.

  Further, in order to maintain the calculation accuracy, as shown in FIG. 3E, normalization processing (renormalization) for expanding the range Range so as to be larger than a predetermined value is performed. For example, in the normalization process, the value of the range Range is doubled and expanded so as to be larger than a predetermined value.

FIG. 4 is a diagram for explaining the normalization process in the encoding process.
For example, as described above, every time a data string to be encoded is encoded in units of 1 bit, the normalization process as shown in FIG. 4 is performed.
Specifically, first, it is determined whether or not the range Range is smaller than a predetermined value (ST1). If the range is smaller than the predetermined value, a code corresponding to the upper 1 bit of the value of the lower limit value Low is determined. Is output (ST2), the range Range and the lower limit value Low are left-shifted to perform double expansion (ST3, ST4), and the process returns to step ST1.
Then, the operations of steps ST1 to ST4 are repeated until the value of the range Range becomes a predetermined value or more.
On the other hand, if the value of the range Range is greater than or equal to a predetermined value in step ST1, the normalization process is terminated.

FIG. 5 is a functional block diagram of the encoding circuit shown in FIG.
For example, as illustrated in FIG. 5, the encoding circuit 27 includes a binarization unit (BINR) 271, a context (CTX) 272, an update unit (UPDT) 273, an arithmetic encoding unit 274, an addition unit 275, and a prediction unit 276. Have
The arithmetic encoding unit 274 corresponds to the arithmetic encoding unit according to the present invention, the adding unit 275 corresponds to the adding unit according to the present invention, and the prediction unit 276 corresponds to the predicting unit according to the present invention.

A binarization unit (BINR) 271 performs binarization processing, for example, an SE input from the quantization circuit 26 or the motion prediction / compensation circuit 39 by using a VLC or a fixed-length code as a string (binary symbol) string (binary symbol). Data) and output as a signal S271.
Specifically, the binarization unit 271 converts the input data into binary data when the input data is a non-binary value (Non Binary Value), for example, a motion vector or a motion coeff.

The context (CTX) 272 is data that is updated with, for example, a statistical value of data encoded by the arithmetic encoding unit 274 and is referred to during arithmetic encoding processing by the arithmetic encoding unit 274.
In the context 272, for example, an index indicating a transition state and an MPS appearance probability pMPS are stored in association with each other. The index of the context 272 is updated by the updating unit 273 based on the signal S274 indicating the processing result of the arithmetic coding unit 274, for example.
Specifically, for example, when the value encoded based on the signal S274 is 1, the update unit 273 updates the index of the context 272 so that the appearance probability of 1 increases, and for example, the encoded value is 0. In this case, the index is updated so that the appearance probability of 0 increases.
By doing so, the index of CTX 273 approaches the appearance probability depending on the input SE.

  FIG. 6 is a diagram for explaining pictures, slices, and macroblocks in image data. FIG. 6A shows that one picture in the image data is composed of a plurality of slices in the horizontal direction, and FIG. 6B shows that one slice is composed of macro blocks. FIG.

For example, the arithmetic coding unit 274 performs arithmetic coding processing on the image data on a block basis, for example, on a macroblock basis, for example, as shown in FIG. 6B, and outputs the encoded data of the processing result as a signal S274.
For example, the arithmetic encoding unit 274 performs arithmetic encoding processing on the input data based on an index indicating the probability of the context 272. In the present embodiment, the arithmetic encoding unit 274 performs an arithmetic encoding process as shown in FIG. 3, for example.
A macro block corresponds to a first block unit according to the present invention, and one slice or one picture corresponds to a second block unit according to the present invention.

The adding unit 275 outputs the signal S274 as encoded data output from the arithmetic encoding unit 274 to a predetermined unit larger than the block unit obtained by the arithmetic encoding process of the arithmetic encoding unit 274, for example, image data in one slice. Additional data corresponding to the index data indicating the degree of compression is added. Additional data corresponds to CabacZeroWord.
As a result of the encoding process performed by the arithmetic encoding unit 274, the additional data is generated when the decoder decodes the encoded data having a high compression rate when the encoded data having a high compression rate is generated. Since processing may not be completed within a predetermined time, for example, as shown in FIG. 6 (a), it is added to the end of each of a plurality of slices (Slice) in one image.
The adding unit 275 reduces the load on the decoder by adding a stuffing byte (Stuffing Byte) having 0x000003 as one unit, for example, as additional data.

  Specifically, the adding unit 275 calculates the index data K as shown in Equation (1), and when the index data K is positive, the bit for (K × 3) bytes is added at the end of one slice. Is added to the encoded data output from the arithmetic encoding unit 274 as additional data.

[Equation 1]
K = Ceil ((Ceil ((3 × BinCountsInNALUnits−3 × 96 × PicSizeInMbs) / 32) −NumBytesInVclNALUnits) / 3) (1)

In Equation (1), BinCountsInNALUnits is data before arithmetic coding processing by the arithmetic coding unit 274, and is, for example, a predetermined unit of data output by the binarizing unit 271, for example, a data length in one picture. .
NumBytesInVclNALUnits is a predetermined unit of data subjected to arithmetic coding processing by the arithmetic coding unit 274, for example, a data length in one picture.
PicSizeInMbs is the number of macroblocks in a predetermined unit, for example, one picture. In this embodiment, it is a fixed value.
Ceil (A) indicates a minimum integer value larger than the value A.

The data length of the additional data added at the end of one slice is 1 because the context 272, the arithmetic encoding unit 274, and the updating unit 273 perform the encoding process while updating the probability model according to the input data. It is indeterminate until the data for the slice is encoded. For this reason, the data length of the additional data added by the adding unit 275 is also uncertain.
For example, in a general rate control circuit, the quantization rate is controlled based on the encoded data generated by the arithmetic encoding unit 274 and the additional data.

  In the present invention, by providing a prediction unit 276, which will be described later, the additional data is predicted in units smaller than one slice, for example, macroblock units. Details will be described below.

  The prediction unit 276 is based on the data length of the data to be subjected to the arithmetic coding processing by the arithmetic coding unit 274 and the data length of the encoded data generated by the arithmetic coding unit 274 in block units, for example, in macroblock units. Thus, the additional data corresponding to the index data in a predetermined unit added by the adding unit 275 is predicted.

  Specifically, the prediction unit 276 is a processing target before the arithmetic coding unit 274 performs arithmetic coding processing while the arithmetic coding unit 274 performs coding processing of data for one slice in units of macroblocks. Based on the data, that is, the data length of the data generated by the binarization unit 271 and the data after the arithmetic encoding process by the arithmetic encoding unit 274, additional data for a predetermined unit, for example, one picture is predicted.

The rate control circuit 32 sets data indicating the degree of quantization of the quantization circuit 26, specifically, a quantization rate, based on the additional data predicted by the prediction unit 276 in block units, for example, macroblock units.
Further, the rate control circuit 32 determines the quantum of the quantization circuit 26 based on the data length of the data obtained by adding the additional data generated by the adding unit 275 to the encoded data in a unit larger than the block unit, for example, one slice unit. Data indicating the degree of quantization, more specifically, a quantization rate is set.
The quantization circuit 26 performs a quantization process according to the set quantization rate.

FIG. 7 is a flowchart for explaining the operation of the encoding circuit 27 shown in FIG. The operation of the encoding circuit 27 will be mainly described with reference to FIG.
In step ST1, the binarization unit 271 converts the SE input from the binarization process, for example, the quantization circuit 26 or the motion prediction / compensation circuit 39, into a sequence of symbols bin (binary symbol) using VLC or a fixed-length code ( Binary data) and output as a signal S271.

  In step ST2, the signal S271 is input to the context 272, and the arithmetic encoding unit 274 performs arithmetic encoding processing according to the probability model set by the context 272 to generate encoded data S274. The updating unit 273 updates the context 272 according to the signal S274.

  In step ST3, the prediction unit 276, in units of blocks, for example, in units of macroblocks, the data length of data to be processed by the arithmetic encoding unit 274, and the data of the encoded data S274 generated by the arithmetic encoding unit 274 Depending on the index data K indicating the degree of compression of image data in units of one picture, for example, in a predetermined unit larger than the block unit by arithmetic coding processing, which is added to the encoded data S274 by the addition unit 275 based on the length Predict additional data.

In step ST4, it is determined whether or not the currently processed current macroblock CurrentMB is in the middle of one slice. Specifically, for example, it is determined whether or not the currently processed current macroblock CurrentMB is smaller than the number of macroblocks in one slice.
When it is determined that the current macroblock CurrentMB processed at present is not in the middle of one slice, specifically, when it is determined that the current macroblock CurrentMB processed at present is not smaller than the number of macroblocks in one slice, That is, when the arithmetic coding unit 274 processes the data for one slice, the adding unit 275 uses the index data as shown in Equation (1) based on the data length before and after the arithmetic coding processing for one slice. K is calculated, and additional data corresponding to the index data K is added to the encoded data for one slice.
The rate control circuit 32 performs quantization of the quantization circuit 26 based on the data length of the data obtained by adding the additional data generated by the adding unit 275 to the encoded data in a unit larger than the block unit, for example, one slice unit. Data indicating the degree, specifically, a quantization rate is set.
The quantization circuit 26 performs a quantization process according to the quantization rate.

  On the other hand, when it is determined in step ST4 that the currently processed current macroblock CurrentMB is in the middle of one slice, specifically, for example, if the currently processed current macroblock CurrentMB is smaller than the number of macroblocks in one slice. If it is determined, the rate control circuit 32 determines the quantization level of the quantization circuit 26 based on the additional data predicted by the prediction unit 276 on a block basis, for example, on a macroblock basis. A rate is set (ST5). In addition, the quantization circuit 26 performs a quantization process according to the set quantization rate.

FIG. 8 is a flowchart showing a specific example of the operation of the arithmetic encoding unit 274 shown in FIG. Next, a specific example of the operation of the arithmetic encoding unit 274 will be described with reference to FIG.
Specifically, the arithmetic encoding unit 274 performs an encoding process with the index ctxIdx of the designated context 272 and the symbol binVal to be encoded as inputs.
For example, codIRange and codILow shown in FIG. 3 represent an appearance probability width (range Range) and a lower limit value Low. The context CTX 272 is managed by an index ctxIdx, and valMPS and pStateIdx indicating an MPS value and a transition state number are assigned to each context.

  In step ST101, the arithmetic encoding unit 274 quantizes the value of codIRange into a value qCodIRangeIdx, and obtains the LPS probability width codIRangeLPS using the specified table value rangeTabLPS. Also change codIRange to the MPS probability range.

  In step ST102, the arithmetic encoding unit 274 branches the process depending on whether or not the input symbol bin is MPS. When the symbol is MPS, the state is changed by the table value transIdxMPS defined in the standard, and pStateIdx is updated (ST103).

On the other hand, in step ST102, when the input symbol bin is LPS, the codILow value and the codIRange value are updated.
Specifically, a value obtained by adding codILow and codIRange is substituted into codILow, and codIRangeLPS is substituted into codIRange (ST104).
In step ST105, it is determined whether or not pStateIdx is 0. If pStateIdx is 0, 1-valMPS is substituted for valMPS (ST106), and the process proceeds to step ST107. Even when pStateIdx is not 0 in step ST105, the process proceeds to step ST107.
In step ST107, pStateIdx is updated.
In step ST108, the arithmetic encoding unit 274 performs a normalization process described later.

FIG. 9 is a flowchart for explaining a specific example of the normalization process shown in FIG. The arithmetic coding unit 274 performs the normalization process in step ST108 shown in FIG. 8 as follows.
As shown in step ST111, the arithmetic encoding unit 274 determines whether or not the value of codIRange is smaller than a predetermined value, for example, 0x100, and determines that the value of codIRange is not smaller than a predetermined value, for example, 0x100 Does not perform normalization processing.

On the other hand, if it is determined in step ST111 that the value of codIRange is smaller than a predetermined value, for example, 0x100, the arithmetic encoding unit 274 determines whether the value of codILow is smaller than a predetermined value, for example, 0x100. (ST112).
When it is determined that the value of codILow is smaller than a predetermined value, for example, 0x100, the arithmetic encoding unit 274 performs a process PutBit (0) for outputting encoded data, which will be described later (ST113), and sets the codIRange value. Multiply by 1/2 and codILow value by 1/2. Specifically, the codIRange value and the codILow value are left-shifted by 1 bit to double each value (ST114), and the process returns to step ST111.

On the other hand, when it is determined in step ST112 that the value of codIRange is not smaller than a predetermined value, for example, 0x100, the arithmetic encoding unit 274 determines whether the value of codILow is a predetermined value, for example, 0x200 or more. (ST115). When it is determined that the value of codILow is not a predetermined value, for example, 0x200 or more, a predetermined value, for example, 0x100 is subtracted from the codILow value, 1 is added to the value of bitsOutstanding (ST116), and the process proceeds to step ST114.
The bitsOutstanding is data provided to prevent carryover during the normalization process, and the bitsOutstanding data is reset to 0 when the encoding process of a predetermined unit is completed.

  On the other hand, if it is determined in step ST115 that the value of codILow is a predetermined value, for example, 0x200 or more, a predetermined value, for example, 0x200 is subtracted from the codILow value (ST117), and encoded data described later is output. Process BitBit (1) is performed (ST118), and the process proceeds to step ST114.

FIG. 10 is a flowchart for explaining the process of PutBit (B) in the encoding process shown in FIG. The operation of PutBit (B) in steps ST113 and ST118 of the arithmetic coding process will be described with reference to FIG.
The arithmetic encoding unit 274 determines whether or not it is the first bit when outputting encoded data. For this determination, for example, firstbitFlag indicating the first bit is used. When firstbitFlag is other than 0, it indicates that it is the first bit, and when firstbitFlag is 0, it indicates that it is other than the first bit.
Specifically, when the firstbitFlag is other than 0 in step ST210, the arithmetic encoding unit 274 assigns 0 to the firstbitFlag and does not output the first bit.
On the other hand, in step ST210, when the firstbitFlag is 0, the arithmetic encoding unit 274 performs WriteBits (B, 1) processing. Here, WriteBits (B, 1) is a process of outputting a value B having a 1-bit length.

  In step ST213, the arithmetic encoding unit 274 determines whether or not bitsOutstanding is greater than 0. If it is determined that the bits Outstanding is greater than 0, the arithmetic coding unit 274 performs WriteBits (1-B, 1) (ST214). 1 is subtracted from bitsOutstanding, and the process returns to step ST213. Here, the process of WriteBits (1-B, 1) is a process of outputting B's complement of 1 bit length.

On the other hand, in Step ST213, when the arithmetic encoding unit 274 determines that bitsOutstanding is 0 or less, the arithmetic encoding unit 274 stops a series of processes.
By performing the operation described above, the arithmetic encoding unit 274 can perform the encoding process without causing carryover.

The arithmetic encoding unit 274 does not output encoded data in order when performing arithmetic encoding processing in units of macroblocks, but when performing arithmetic encoding processing using bitsOutstanding as described above, bitsOutstanding. The encoded data is output according to the value of.
For this reason, it is preferable that the prediction unit 276 performs prediction in consideration of the bitsOutstanding.
Hereinafter, a specific example of the prediction unit 276 will be described in detail.

  The prediction unit 276 calculates the data length of the data generated by the binarization unit 271, that is, the data length BitCounts of the data to be processed by the arithmetic encoding unit 274 until, for example, the macroblock currently being processed within one slice. To do.

Also, the prediction unit 276 calculates the data length NumBytes of the encoded data generated up to the macro block currently being processed.
Specifically, the prediction unit 276, as the data length NumBytes, for example, as shown in Formula (2), the data length GenBits of the encoded data generated up to the currently processed macroblock (also referred to as CurrentMB), Calculate by adding the value of bitsOutstanding at the time.

[Equation 2]
NumBytes = (GenBits + bitsOutstanding) / 8 (2)

As described above, bitsOutstanding is added because bitsOutstanding originally counts the amount of bits output to the buffer at that time in order to avoid carryover during normalization processing.
By the calculation shown in Expression (2), the prediction unit 276 can calculate the data length NumBytes of the encoded data generated up to the macro block currently being processed with higher accuracy.

  Since GenBits is a bit that is actually encoded and output to the buffer, it includes an emulation prevention byte.

  In addition, the prediction unit 276 processes the data for one picture in units of macroblocks, for example, as shown in Equation (3), the data length of the data after binarization by the binarization unit 271 for one picture The estimated value EstimateBinCountsInNALUnits is calculated using the data length CurrentBinarizationNum of the binarized data up to CurrentMB, the number of macroblocks PicSizeInMbs of one picture, and the data MbNo indicating the currently processed macroblock.

[Equation 3]
EstimateBinCountsInNALUnits = CurrentBinarizationNum x (PicSizeInMbs / MbNo) (3)

Further, the prediction unit 276 processes one picture data in units of macroblocks, for example, as shown in Equation (4),
Data length prediction value EstimateNumByteInVclNALUnits of encoded data for one picture, data length GenBits, bitsOutstanding value of data output by arithmetic encoding unit 274 through arithmetic encoding for one picture, and macro of one picture Calculation is performed using the number of blocks PicSizeInMbs and the data MbNo indicating the currently processed macroblock.
MbNo is a macro block currently being processed. Since this is added for each macro block, it is the sum of the macro blocks encoded so far.

[Equation 4]
EstimateNumByteInVclNALUnits = (GenBits + bitsOutstanding) x (PicSizeInMbs / MbNo) / 8 ... (4)

  The prediction unit 276 predicts index data K indicating the degree of compression for one picture by substituting the predicted values EstimateBinCountsInNALUnits and EstimateNumByteInVclNALUnits into BinCountsInNALUnits and NumBytesInVclNALUnits shown in Equation (1).

  Specifically, the prediction unit 276 predicts the index data K by performing a calculation as shown in Equation (5).

[Equation 5]
K = Ceil ((Ceil ((3 × EstimateBinCountsInNALUnits−3 × 96 × PicSizeInMbs) / 32) −EstimateNumByteInVclNALUnits) / 3) (5)

Further, the prediction unit 276 predicts the data length of the additional data according to the predicted index data K.
The rate control circuit 32 controls the quantization rate in units of macroblocks according to the data length of the additional data corresponding to the predicted index data K.

As described above, the prediction unit 276 can predict the additional data generated by the addition unit 275 in units of one macroblock. Also, the index data K and additional data predicted by the prediction unit 276 approach the index data K and additional data generated by the addition unit 275 as the arithmetic coding process proceeds and approaches the last macroblock of one picture.
Since the operation of each component is the same as that of the flowchart shown in FIG.
The difference is that the prediction unit 276 predicts additional data based on Equations (2) to (5) in consideration of bitsOutstanding. Thereby, additional data can be predicted with higher accuracy.

  As described above, the arithmetic coding unit 274 that performs arithmetic coding processing on image data in units of macro blocks to generate coded data, and the coded data generated by the arithmetic coding unit 274 are added to the arithmetic coding unit 274. An addition unit 275 for adding additional data corresponding to the index data K indicating the degree of compression of the image data in a predetermined unit larger than the macroblock unit by the arithmetic coding process, and the arithmetic coding unit 274 in the macroblock unit. Addition according to the index data K in a predetermined unit added by the adding unit 275 based on the data length of the data to be subjected to arithmetic coding processing by and the data length of the encoded data generated by the arithmetic coding unit 274 A prediction unit 276 that predicts data, and a rate control circuit 3 that performs predetermined processing according to additional data predicted by the prediction unit 276 in units of macroblocks And since there is provided a quantization circuit 26, in the encoding process, we can predict the additional data in units smaller than the predetermined unit of additional data of the image data.

  Also, the bit outstanding data bitsOutstanding related to the normalization processing by the arithmetic encoding unit 274, the data length of the encoded data up to the block that the arithmetic encoding unit 274 has arithmetically processed among one picture of the image data, and 1 Since the additional data corresponding to the index data in a predetermined unit is predicted based on the data length of the data before arithmetic coding processing in the picture, the index data and the additional data can be predicted with higher accuracy.

Note that the present invention is not limited to this embodiment, and any suitable modification can be made.
In the encoding apparatus according to the present embodiment, the functions according to the present invention are realized by hardware, but the present invention is not limited to this form. For example, the function according to the present invention may be realized by executing a program having the function according to the present invention described above by a computer (data processing apparatus).

It is a conceptual diagram of the communication system 1 which employ | adopted the encoding apparatus which concerns on one Embodiment of this invention. It is a whole block diagram of the encoding apparatus 2 shown in FIG. It is a figure for demonstrating arithmetic coding processing. It is a figure for demonstrating the normalization process in an encoding process. FIG. 3 is a functional block diagram of the encoding circuit shown in FIG. 2. It is a figure for demonstrating a picture, a slice, and a macroblock in image data. (A) is a diagram showing that one picture in the image data is composed of a plurality of slices in the horizontal direction, and (b) shows that one slice is composed of macroblocks. FIG. 3 is a flowchart for explaining the operation of the encoding circuit shown in FIG. 2. 6 is a flowchart illustrating a specific example of the operation of the arithmetic encoding unit illustrated in FIG. 5. It is a flowchart for demonstrating one specific example of the normalization process shown in FIG. 10 is a flowchart for explaining a process of PutBit (B) of the encoding process shown in FIG. 9.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 ... Communication system, 2 ... Encoding apparatus (data processing apparatus: Computer), 3 ... Decoding apparatus, 22 ... A / D conversion circuit, 23 ... Screen rearrangement circuit, 24 ... Arithmetic circuit, 25 ... Orthogonal transformation circuit, 26 DESCRIPTION OF SYMBOLS ... Quantization circuit, 27 ... Coding circuit, 28 ... Buffer, 29 ... Inverse quantization circuit, 30 ... Inverse orthogonal transformation circuit, 32 ... Rate control circuit, 34 ... Frame memory, 38 ... Deblock filter, 39 ... Motion prediction Compensation circuit, 271 ... binarization unit (BINR), 272 ... context (CTX), 273 ... update unit (UPDT), 274 ... arithmetic coding unit, 275 ... addition unit, 276 ... prediction unit, Range ... range, Low: Lower limit.

Claims (6)

Arithmetic coding means for performing coded coding of image data on a macroblock basis to generate coded data;
Wherein the arithmetic coding unit generated encoded data, the degree of the generated I by the arithmetic coding process in the arithmetic encoding means, of the image data in the slice or picture unit including a plurality of macro blocks compressed adding means for adding additional data corresponding to the index data indicating,
In the macro-block unit, the data length of the arithmetic coding unit target data arithmetic coding process by, and, on the basis of the data length of the coded data arithmetic coding means is generated, the additional means for adding A prediction means for predicting additional data;
And a processing unit that performs predetermined processing according to the additional data predicted by the prediction unit in units of the macroblock . The arithmetic encoding means includes a normalization process for multiplying the data to be processed by a predetermined number when performing the arithmetic encoding process,
The prediction means includes bit-outstanding data related to the normalization processing by the arithmetic coding means, and encoded data up to a macroblock that has been arithmetically coded by the arithmetic coding means in the picture of the image data. The encoding apparatus according to claim 1, wherein additional data corresponding to the index data is predicted based on a data length and a data length of the data before the arithmetic encoding process in the one picture. The processing means performs quantization processing based on data indicating a set degree of quantization of the image data, and generates quantization target processing data of the arithmetic encoding means;
In the macro-block basis, on the basis of the said additional data predicting means predicts the encoding of claim 1 including the quantized data setting means for setting data indicating the degree of quantization of the quantization means apparatus. Motion vector generation means for generating a motion vector based on a difference between image data to be subjected to motion compensation and reference image data referred to in the motion compensation in units of a plurality of macro blocks of different sizes;
First processing means for sequentially performing orthogonal transformation processing and quantization processing on the difference between the image data to be subjected to motion compensation and the predicted image data;
Second processing means for generating reference data by sequentially performing inverse quantization processing and inverse orthogonal transform processing on the data generated by the first processing means;
Third processing means for generating the predicted image data based on the motion vector and the reference image data;
Quantization means for performing quantization processing based on data indicating a set degree of quantization and generating data to be processed by the encoding means;
Arithmetic encoding means for generating encoded data by performing arithmetic encoding on the data generated by the quantization means and the motion vector in units of the macroblocks ;
Wherein the arithmetic coding unit generated encoded data, the degree of the generated I by the arithmetic coding process in the arithmetic encoding means, of the image data in the slice or picture unit including a plurality of macro blocks compressed Adding means for adding additional data according to the index data indicating
In the macro-block unit, the data length of the arithmetic coding unit target data arithmetic coding process by, and, on the basis of the data length of the coded data arithmetic coding means is generated, the additional means for adding prediction means for predicting the additional data corresponding to the index data,
And a quantized data setting unit configured to set data indicating a degree of quantization of the quantizing unit based on the additional data predicted by the predicting unit for each macroblock . A program to be executed by an information processing apparatus having an encoding unit, an adding unit, a predicting unit, and a processing unit ,
A first procedure in which the encoding means performs an arithmetic encoding process on image data in units of macroblocks to generate encoded data;
The adding unit compresses the image data in units of slices or pictures including a plurality of macroblocks generated by the arithmetic encoding process of the arithmetic encoding unit in the encoded data generated by the arithmetic encoding unit. A second procedure for adding additional data according to the index data indicating the degree of
The prediction means is based on the data length of the data subject to arithmetic coding processing in the first procedure and the data length of the coded data generated in the first procedure in units of the macroblocks. a third step for predicting the additional data to be added in the second step,
A fourth procedure in which the processing means performs a predetermined process according to the additional data predicted in the third step in units of the macroblock ;
A program for causing the information processing apparatus to execute. An encoding processing method performed by an information processing apparatus including an encoding unit, an adding unit, a prediction unit, and a processing unit ,
It said encoding means, a first step of generating encoded data subjected to arithmetic encoding processing image data in units of macroblocks,
The adding unit compresses the image data in units of slices or pictures including a plurality of macroblocks generated by the arithmetic encoding process of the arithmetic encoding unit in the encoded data generated by the arithmetic encoding unit. A second step of adding additional data according to the index data indicating the degree of
The prediction means, in the macro-block unit, the data length of the target data of the arithmetic encoding processing in the first step, and, based on the data length of the coded data generated in said first step, said a third step of predicting the additional data to be added in the second step,
It said processing means, in the macro block, and a fourth step of performing a predetermined process corresponding to the additional data predicted in the third step,
An encoding processing method comprising:

Images