JP3254070B2 - Image coding method - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a coding method for high-efficiency image coding, and more particularly, to a method in which the amount of information for each single frame or a plurality of frames of an image is quantized into predetermined plural bits (for example, bytes). The present invention relates to an image coding method which is restricted to take a value.

[0002]

2. Description of the Related Art As a conventional image encoding method, when an arbitrary encoding bit rate and an arbitrary frame rate are given, the information amount of an image per single or a plurality of frames is converted into a plurality of bits as a quantization unit. No method has been reported in which the setting is made to be an integral multiple of, and the information amount of any single or multiple frames is made equal.

[0003]

In the conventional coding method for high-efficiency image coding, the amount of information in units of a plurality of frames is generally not always uniform. Therefore, when editing a bit stream (binary coded data) obtained by encoding, if the temporal order of images is changed due to the editing, the amount of generated information is temporally biased. For example, as a result of editing,
The bit rate in the first half of the image sequence is abnormally high or conversely low. As a result, when trying to decode the edited bit stream, the read buffer prepared for normal playback may run out of capacity, causing a buffer overflow or conversely making playback impossible due to underflow. Was.

In the case where the buffer level for each frame of an image is written in the bit stream, it is necessary to recalculate and rewrite the buffer level after editing in the conventional method, which requires extra processing. there were.・ Reference: Hiroshi Yasuda, "International Standards for Multimedia Coding", Chapter 6, Maruzen. An object of the present invention is to solve the above problems and to provide an image encoding method capable of creating a bit stream that can be easily edited.

[0005]

Hereinafter, the present invention will be described in detail. Here, a group of images of a plurality of frames as a unit of editing is called a GOP (Group Of Pictures). It is assumed that the number of bits given to one frame of an image is quantized in byte units. In this example, a byte (8 bits) unit is used, but any other number of bits can be similarly handled. Set any coding bit rate to R (bit
/ Sec), and the frame rate (the number of frames per second) is P (frame / sec). Also, GOP
Let N be the number of frames included in.

First, the number of bytes G (bytes) that can be given to a GOP is determined. The number of bits a given to the GOP is obtained with integer precision using the following equation. a = RN / P (1) Here, “/” represents a truncation operation. Next is 1 byte 8
Since it is a bit, a in equation (1) is divided by 8, and the remainder is truncated.

G = a / 8 (2) Further, the byte of Expression (2) is converted into bits, and the encoding bit rate is recalculated. As a result, an encoding bit rate Re (bit / sec) corrected for editing is obtained.

Re = 8 GP / N (3) The bit rate deviation R−Re (bit) is handled by the buffer of the data multiplexing unit. That is, a shift in units of 1 bit is dealt with by inserting dummy data in the multiplexing portion.
When encoding an image, the target bit rate is set to be smaller than Re (for example, 0.9 Re), and the shortfall is inserted by inserting dummy bytes for making up the specified amount.
Adjust to e.

Normally, the amount of information generated for each GOP does not exceed 8 G by setting the target bit rate slightly lower than Re. However, the amount of generated information may exceed 8G. In that case, the excess is recorded, the target number of bits for the next GOP is subtracted by the excess, and dummy bits are inserted so that the information amount of 16 G is obtained for 2 GOPs. At this time, at the time of editing, the editing unit of the bit stream is partially processed as 2 GOPs.

[0010]

According to the present invention, since the amount of information generated for each GOP takes a constant value in byte units, buffer overflow,
Underflow does not occur, and the bit stream in GOP units can be easily edited.

Also, when the buffer level for each frame of an image is written in the bit stream,
Since there is no change in the buffer level for each OP, there is no need to recalculate and rewrite the buffer level after editing.

[0012]

Embodiments of the present invention will be described below. FIG. 1 shows a block diagram of an image encoding method according to an embodiment of the present invention. The given encoding bit rate, frame rate, and encoding parameter 1 of the number of frames included in the GOP are input to the encoding parameter setting unit 2. The encoding parameter setting unit 2 performs an encoding parameter setting process in accordance with the flowchart shown in FIG. 2 and outputs a modified parameter 3.

[0013] In the encoding unit 4, the modified parameter 3
, And the generated information amount 6 is measured for each frame. Based on the generated information amount 6, the bit number inspection unit 7 measures how much the generated information amount is insufficient for the number of bits in the editing unit among the corrected parameters. The bit number checking unit 7 outputs the number of bits of the shortage information 8, and inputs it to the dummy bit inserting unit 9. From the dummy bit insertion unit 9, a dummy bit 10 for each editing unit over a single frame or a plurality of frames is converted into a bit stream 5 obtained by encoding by the encoding unit 4.
Inserted for

By inserting dummy bits, a video bit stream 11 having a uniform information amount in units of a plurality of frames.
Is calculated as the amount of information in a certain period of time, and the dummy bit insertion unit 12 in the multiplexing unit compares it with the original encoding bit rate of the encoding parameters. The dummy bits 13 are added to the video bit stream 11 and output as a multiplexed bit stream 14.

FIG. 2 shows a flowchart of a coding parameter setting section in an image coding method according to an embodiment of the present invention. As shown in FIG. 2, the encoding parameter setting unit 2 calculates the GOP from the encoding parameter 1 (the encoding bit rate, the frame rate, and the number of frames included in the GOP).
The number of bits per GOP (a = RN / P) is calculated by a calculating unit 21 of the number of bits per GOP, and G = a / 8 is calculated by the calculating unit 22 of the number of bytes per GOP. Truncated. Based on this value, a modified coding bit rate Re = 8GP / N is calculated,
A modified coding parameter 3 is obtained.

FIG. 3 shows an example of a bit stream encoded according to an embodiment of the present invention. The number of frames included in one GOP is N. Encoding from the first frame to the Nth frame with the modified encoding bit rate Re, and dummy bits 10
Insert Then, when multiplexing coded data, DCT coefficients, picture headers, etc., dummy bits 13 are inserted by comparing with the original coding bit rate.

FIG. 4 shows an embodiment corresponding to the second aspect of the present invention, and is an explanatory diagram of processing when the amount of generated information exceeds a prescribed amount of information. In FIG. 4, reference numeral 31 denotes processing of the first GOP by the block diagram shown in FIG.
Indicates the processing of the second GOP by the block diagram shown in FIG. 1, and 33 indicates the modification of the encoding parameter.

As described above, the target bit rate is set slightly smaller than the corrected encoding bit rate Re. However, in processing 31 of a certain GOP (first GOP), the amount of generated information is 8 G (G is given to the GOP). Number of bytes that can be
If it is determined by the bit number checking unit 7 that the number of bits exceeds the limit, the excess is recorded, and the target parameter number for the next GOP (second GOP) is subtracted by the excess to correct 33 the encoding parameter. In the process 32 of the second GOP, the process according to the block diagram shown in FIG. 1 is executed by using the modified encoding parameter 1 ', and the bit number check unit 7' adjusts the information amount to 16G in 2 GOPs. To determine the number of bits to perform. Dummy bits are inserted by the number of bits. At the time of editing, the editing unit of the bit stream is partially processed as a 2 GOP obtained by combining the first GOP and the second GOP.

FIG. 5 shows an example of the configuration of the entire system to which the present invention is applied. The encoding device 41 applies the image encoding method according to the present invention to the original video input 40,
A multiplexed bit stream 14 is generated in which the amount of information generated for each GOP has a constant value in byte units. As a result, the editing device 42 can easily edit the bit stream in GOP units. By decoding the edited bit stream 43 by the decoding device 44, an edited video reproduction signal 45 is output. Since the editing device 42 and the decoding device 44 may be the same as those in the prior art, detailed description is omitted here.

[0020]

As described above, according to the present invention,
Even if the bit stream is edited, a bit stream whose reproduction is guaranteed can be realized. GOP after editing
There is no need to recalculate and rewrite the buffer level for each, and the bit stream can be easily edited.

[Brief description of the drawings]

FIG. 1 is a block diagram of an image encoding method according to an embodiment of the present invention.

FIG. 2 is a flowchart of an encoding parameter setting unit in an image encoding method according to an embodiment of the present invention.

FIG. 3 is a diagram illustrating an example of a bit stream encoded according to an embodiment of the present invention.

FIG. 4 is an explanatory diagram of a process when an amount of generated information exceeds a specified amount of information in the embodiment of the present invention.

FIG. 5 is a diagram showing a configuration example of an entire system to which the present invention is applied.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Coding parameter 2 Coding parameter setting part 3 Modified parameter 4 Coding part 5 Bit stream 6 Generated information amount (bit) 7 Bit number inspection part 8 Shortage information 9 Dummy bit insertion part 10 Dummy bit 11 Video bit stream 12 Dummy bit insertion unit in multiplexing unit 13 Dummy bit 14 Multiplexed bit stream

Continuation of the front page (56) References JP-A-63-279686 (JP, A) JP-A-4-167868 (JP, A) JP-A-7-184196 (JP, A) Yasuo Katayama, "MPEG2 quantization and Coding Control, ”Technical Report of the Institute of Television Engineers of Japan, Japan, October 15, 1992, Vol. 16, No. 61, p. 43-48 (58) Fields surveyed (Int. Cl. ⁷ , DB name) H04N ^7/ 24-7/68

Claims (2)

(57) [Claims] 1. An image coding method in which the amount of information for each single or multiple frames of an image is limited to take a value quantized in predetermined multiple bits. When an arbitrary frame rate is given, the amount of information of an image per single or multiple frames is set to be an integral multiple of the multiple bits as a quantization unit, and Reset the modified coding bit rate closest to the given coding bit rate so that the amount of information is also equal,
Bit stream that spans one or more frames during encoding
About the shortage of generated information for each editing unit of the ream
An image coding method, comprising inserting dummy bits and processing a deficiency with respect to an actual coding bit rate by inserting dummy bits when multiplexing coding information. 2. The image encoding method according to claim 1, wherein when an amount of generated information of an image per one or a plurality of frames encoded based on the modified encoding bit rate exceeds a prescribed information amount, The excess is recorded, and the target number of bits for the next single or multiple frames is subtracted by the excess, and dummy bits are inserted so that the generated information amount of the combined image is twice the specified information amount Image encoding method.