JPH06113289A - Scanning selection method and device for picture coding - Google Patents

Abstract

PURPOSE:To select the scanning method minimizing a code quantity when picture data are quantized, scanned and coded. CONSTITUTION:A quantized signal 53 is applied to scanners 241 connected in parallel, the signal is scanned by various scanning methods, number of zeros in the obtained scanning data is accumulated by each zero number accumulator 245 till a value not zero appears, an output of a scanner 241 whose number of zeros is maximum is selected by a maximum selector 246 and fed to a coder 25. Thus, when scanning data by the scanning method maximizing number of zeros are coded, since the code quantity is minimum, efficient coding is attained. Since the zero number accumulator 245 accumulates only number of zeros, the high speed operation is attained and the configuration is simple.

Description

Detailed Description of the Invention

[0001]

FIELD OF THE INVENTION The present invention relates to a scan selection method and apparatus in image coding. Specifically, it is an object of the present invention to provide a scanning selection method and device in image coding that reduces the value of large prediction error data that tends to occur when motion-compensating and predictively coding an image signal for transmission.

[0002]

2. Description of the Related Art In videophones and video conferences,
The moving image signal to be transmitted has a huge amount of information. Therefore, conventionally, various methods have been used to efficiently encode a moving image signal by encoding the moving image signal. By utilizing the correlation between the pictures (frames or fields) of the image signal, the one used for this high efficiency encoding,
There is inter-picture predictive coding that predicts the current picture from the previous or subsequent picture. The difference value for each pixel between the image signal of the current picture and the image signal of the immediately preceding picture is obtained as prediction error data, and only the obtained prediction error data is encoded and transmitted. This allows
The information amount of the image to be encoded and transmitted is reduced.

The circuit configuration of the entire conventional image encoding apparatus will be described with reference to FIG.

Reference numeral 50 denotes an image signal input terminal. An input image signal 51 input from the input terminal 50 subtracts a prediction signal 60 from a subtractor 20 to obtain a prediction error signal 52. The prediction error signal 52 is orthogonally transformed, for example, discrete cosine transformed by the transformer 22, quantized by the quantizer 23, and the quantized signal 53 is output, which is the scan selector 24.
The scan type is selected by B, and the Huffman encoder 25 that outputs the variable length code outputs the two-dimensional variable length code data 54 of run and level. Further, the quantized signal 53 is inversely quantized by the inverse quantizer 31, and the inverse converter 32 performs inverse conversion of the conversion of the converter 22 to obtain a quantized error signal 55. The adder 34 adds the quantized error signal 55 and the prediction signal 60 to obtain a decoded signal 56.

Decoded signal 56 is applied and stored in frame memory 38. An image signal of a motion-compensated block from a picture stored in the frame memory 38 is generated by the motion compensation circuit 42 in accordance with an instruction of a vector signal 62 which is an output of a motion vector detector 46 that detects a motion vector from an input image signal 51. The motion-compensated signal 64 is read out and is applied to the terminal a of the changeover switch 45.

When the terminal a is selected by the changeover switch 45, the motion-compensated signal 64 is selected and is output as the prediction signal 60. When the terminal b is selected, “0” is output as the prediction signal 60. The changeover switch 45 is changed over by a control signal from the outside (not shown).

In the scan selector 24B, the 8 ×
8 pixels are obliquely zigzag-scanned in the order shown by the numbers in the box of FIG. 3, or 8 × 8 pixels shown in FIG. 4 are zigzag-scanned in the vertical direction shown by the numbers in the box of FIG. The vertical zigzag scanning is performed or the one with a smaller code amount is selected, and either scanning is performed and the output is applied to the Huffman encoder 25.

In the Huffman encoder 25, the scan selector 2
The signal from 4B catches the continuous length (run) of "0, 0, ..., 0" and the value (level) of the non-zero coefficient which is not "0" as an event, and indicates this and the end of this event. The end-of-block event and the escape event indicating the occurrence of a rare combination of run and level are converted into a variable length code.

Regardless of whether it is for communication or storage media, a moving image often has a basic structure of an interlaced field and a frame composed of two fields. Generally, when encoding is performed in the moving region of an interlaced image, the high frequency component of the discrete cosine transform coefficient in the vertical direction appears strongly not only in the input image but also in the prediction error signal. For this reason, if the encoding is performed by the diagonal zigzag scanning in the order shown by the numbers in the box in FIG. 3 of the normal discrete cosine transform coefficient, the encoding efficiency may be reduced.

FIG. 5 shows a circuit configuration of the conventional scan selector 24B. The quantized signal 53 for preventing such a reduction in coding efficiency is a plurality of scanners 241.
-1 to 241-n perform scan conversion by various scanning methods such as the oblique zigzag scanning and vertical zigzag scanning shown in FIGS. 3 and 4, receive the output, and receive the respective code amount generators 242-1 to 242-1. A minimum value selector 243 that generates a code amount in 242-n and selects a scanning method that obtains the smallest code amount among the respective code amounts selects the optimum scanning method, and the data by the scanning method is Huffman coded. It was applied to the digitizer 25, Huffman coded, and the variable length code data 54 was output.

[0011]

Since a plurality of sets of scanners 241 shown in FIG. 5 and a scanning method that minimizes the code amount are selected, the structure is complicated, which makes it difficult to operate at high speed and increases the cost. There is a problem to be solved that a plurality of code amount generators 242 are required.

[0012]

In an apparatus including a scan selector for selecting a scanning method for each luminance block for motion-compensating and predictively encoding an image signal,

A plurality of scanners for scanning a plurality of types to obtain respective scan data;

A zero count accumulator for accumulating the number of zeros after the end of block event appearing in the scan data to obtain a plurality of accumulation results,

A scan selector including a maximum value selector for selecting the scan data showing the maximum value among a plurality of accumulation results was used.

[0016]

In general, when scanning data quantized by two-dimensional orthogonal transformation, non-zero data is often concentrated in part. Therefore, in general, the larger the number of consecutive zeros in the scan data, especially the larger the number of zeros after the end of block event, the smaller the code amount in the case of encoding. Using the scanning method that maximizes the number of zeros after the event, it is possible to obtain almost the minimum code amount. To detect the scanning method that maximizes the number of zeros, it is possible to use a zero-number accumulator that accumulates the number of zeros and that can operate at high speed. It was not necessary to use a code amount generator having a simple structure, and a device capable of high-speed operation could be realized at low cost.

[0017]

FIG. 1 shows the circuit configuration of an embodiment of the scan selector 24 according to the present invention. The scan selector 24 is used in place of the scan selector 24B shown in FIGS. 2 and 5, and the components corresponding to those in FIG. 5 are designated by the same reference numerals.

Quantized signal 53 from quantizer 23
Are simultaneously applied to a plurality of scanners 241-1 to 241-n, and each scanner performs various kinds of scanning such as oblique zigzag scanning and vertical zigzag scanning shown in FIGS. 3 and 4, and each scan data is zero. Number accumulator 245-1 to 245
In -n, the number of zeros until one non-zero (a value other than zero) appears in the order opposite to the scanning arranged in one dimension is output, and the number is output. The maximum value selector 246 that has received each zero number accumulated value outputs the scan data of the scanner 241 that shows the maximum value of each zero number accumulated value, and the Huffman encoder 25 (FIG. 2). Apply to.

The scanning data to be encoded by the Huffman encoder 25 is obtained by scanning the quantized signal 53 in a scanning order that can be encoded with a substantially minimum code amount, so the Huffman encoder 25 is used. The code amount of the variable-length code data 54, which is the output of, is almost the minimum.

N of each scanner 241 and zero-number accumulator 245
When non-zero data is already present in the n-1 set of zero-number accumulators 245 when the sets are operating in parallel in parallel, the remaining zero-number accumulator of the remaining one zero-number accumulator 245 is accumulated. If the accumulated value already exceeds the other accumulated values even if the calculated value has not been obtained yet, this is the optimum scanning method (scan with zero accumulated value showing the maximum value). Method), the scanning data of the scanning device 241 which sends the scanning data to the remaining one zero-number accumulator 245 is selected by the maximum value selector 246, and the Huffman encoder 2 is selected.
You may output to 4.

Further, the scan data are arranged one-dimensionally and the number of zeros is accumulated in the scanning order. When a non-zero appears, the zero number accumulator 245 is reset, and when the scanning is completed,
Zero count accumulator 245 may be configured to output the number of zeros after the ob-block event.

Since the function of the zero-number accumulator 245 is a simple one that simply accumulates the number of zeros until non-zero data appears, the structure may be simple and the accumulation speed may be sufficiently high. Many. Therefore, a small number of zero-number accumulators 245 for the plurality of scanners 241 may be used in time division.

[0023]

As is apparent from the above description, according to the present invention, in order to check the code amount after encoding, it is possible to realize a simple configuration without using a plurality of code amount generators having a complicated configuration. By using the zero-number accumulator, it is possible to obtain scan data by the scanning method that minimizes the code amount, and thus it is possible to inexpensively realize a device that can operate at high speed. Therefore, the effect of the present invention is extremely large.

[Brief description of drawings]

FIG. 1 is a circuit configuration diagram showing an embodiment of a scan selector according to the present invention.

FIG. 2 is a circuit configuration diagram showing a circuit configuration of an entire conventional image encoding device.

FIG. 3 is a processing sequence diagram showing a processing order of diagonal zigzag scanning of 8 × 8 pixels included in one conventional block.

FIG. 4 is a processing sequence diagram showing a processing order of vertical zigzag scanning of 8 × 8 pixels included in one conventional block.

5 is a circuit configuration diagram of a scan selector which is a component of FIG. 2 showing a conventional example.

[Explanation of symbols]

 20 Subtractor 22 Converter 23 Quantizer 24, 24B Scan Selector 25 Huffman Encoder 31 Inverse Quantizer 32 Inverse Transformer 34 Adder 38 Frame Memory 42 Motion Compensation Circuit 45 Changeover Switch 46 Motion Vector Detector 50 Input terminal 51 Input image signal 52 Prediction error signal 53 Quantized signal 54 Variable length code data 55 Quantization error signal 56 Decoded signal 60 Prediction signal 62 Vector signal 64 Motion compensated signal 241-1 to 241-n Scan 242-1 to 242-n Code amount generator 243 Minimum value selector 245-1 to 245-n Zero number accumulator 246 Maximum value selector

 ─────────────────────────────────────────────────── ─── Continuation of the front page (72) Inventor Koichi Oyama 7-5 Minami-Aoyama, Minato-ku, Tokyo Column 6th floor Minami-Aoyama GC Technology Co., Ltd.

Claims (3)

[Claims] 1. A scan selection method (24) in image coding, wherein an image signal is two-dimensionally orthogonally transformed and quantized and selected from a plurality of scan methods for each block, wherein each of the scan methods (241-1 to 241). -N), the number of zeros after the end of block event in the scan data is accumulated (245-1 to 245-n), and the number of zeros is maximized. A scanning selection method in image encoding, wherein the scanned data is encoded (246). 2. A scan selection apparatus for image coding, comprising scan selection means (24) for selecting from a plurality of scanning methods for each block by two-dimensionally orthogonally transforming and quantizing an image signal. Scanning means (241-1 to 241-n) for obtaining respective scan data by each of the above-mentioned scanning methods, and an end-of-second appearing in each of the scan data.
Zero number accumulation means (245-1 to 245-1) for accumulating the number of zeros after the block event to obtain respective accumulation results.
245-n), and a maximum value selection means (246) for outputting to select and encode the scan data indicating the maximum value of the respective accumulation results obtained by the zero number accumulation means. Scan selection device in image encoding. 3. When the maximum value selecting means selects the scan data indicating the maximum value among the respective accumulation results, the accumulated data excluding the remaining one of the respective accumulation results has already been removed. When the maximum value of the calculation result is obtained and the remaining one accumulation result is not obtained, but the accumulated value at the present time has already exceeded the maximum value of the accumulation results excluding the remaining one. 3. The scanning selection device in image coding according to claim 2, wherein the scanning selection device outputs the scanning data for which the remaining one accumulation result is to be selected and encoded.