JPH08149462A - Image packet coder and encoder therefor - Google Patents

Abstract

PURPOSE: To obtain image packet coder and encoder capable of instantly restoring a reception error when the abolition of an image is generated. CONSTITUTION: When a screen is divided by the screen division circuit 2 of an image packet coder 1, the serial number is imparted for every small screen. This number and the small screen coded in a coding circuit 3 are made to correspond and are assembled to a packet in a packet multiplex circuit 4 and the packet is transmitted. In an image packet decoder 10, the number of a small abolished screen can easily be recognized by detecting the interruption of the continuity of the number for every small screen. By returning this small screen number to a transmission side 1 and performing an intra-frame coding for the small screen of this number on the transmission side 1, a reception error can instantly be restored.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image packet coding device and a decoding device, and more particularly to an image packet coding device and a decoding device used for a video conference.

[0002]

2. Description of the Related Art As a method of transmitting a compressed image by subjecting an input image signal to high-efficiency encoding such as inter-frame encoding, a circuit switched network such as an ISDN line which can always maintain a constant transmission bit rate is used. Method and LA
There is a method of using a packet switching network that has a variable bit rate depending on the availability of a line such as N.

High-efficiency coding such as interframe coding is a method of removing temporal redundancy by taking a difference between consecutive image frames, and therefore a lot of much information is input when the motion of an input image is large. However, there is a characteristic that the amount of generated information becomes smaller as the movement becomes smaller.

By the way, when dynamically changing information is transmitted using a fixed rate line such as an ISDN line, it is necessary to control the image quality according to the rate. That is, if the motion of the input image is large, the frame is thinned out or the image quality is reduced to limit the amount of generated information. Conversely, if the motion is small, the image quality is increased or the worst stack bit is inserted to transmit information. The amount is kept constant. Therefore,
There are drawbacks that the image quality changes depending on the movement of the input image and unnecessary information needs to be sent.

On the other hand, since the packet line can control the amount of information that can be transmitted according to the amount of generated information, when the movement of the input image is large, it is possible to increase the transmission packet size or shorten the transmission interval, and vice versa. When the movement is small, the transmission packet size can be reduced or the transmission interval can be lengthened. Therefore, it is possible to maintain the image quality to some extent, and it is possible to almost eliminate the drawbacks when the fixed rate line is used.

[0006]

However, while a fixed rate line is always provided with a constant rate transmission line, a packet line cannot provide a desired rate due to the load state of the line or, in the worst case, on a network. There is a possibility of packet error or discard. In this case, there is a problem that an error occurs in the received image, and the error remains between frames in the interframe coding method and the error cannot be recovered.

By the way, the following prior art is disclosed as a measure for minimizing the reduction of the amount of information due to the discard of the image.

First, in Japanese Patent Laid-Open No. 4-2274, an image signal transmitting apparatus configured to retransmit the packet in which the error is detected to the receiving side when the receiving side detects an error in the image data in the packet line. It is disclosed.

Further, in Japanese Patent Laid-Open No. 2-254886, the image component is divided into an average value component and a component other than the average value component and separately encoded, and when the image is discarded, the components other than the average value component are An image packet encoder and decoder configured to be discarded is disclosed.

Further, Japanese Laid-Open Patent Publication No. 3-234139 discloses an image packet multiplexing system configured to divide images into classes according to their importance and discard images of low importance.

Further, in Japanese Patent Laid-Open No. 15791/1990, a variable structure in which an image is hierarchically divided into a part having a large influence on the image quality and a part having a small influence on the image quality and only the part having a small influence on the image quality is discarded A rate image coding scheme is disclosed.

However, these prior arts do not disclose a technique for immediately recovering an error when an image is discarded and an error occurs in a received image.

Therefore, an object of the present invention is to provide an image packet coding apparatus and a decoding apparatus capable of immediately recovering a reception error when an image is discarded.

[0014]

In order to solve the above-mentioned problems, the present invention divides an input image signal of one screen into a plurality of image signals of small screens, and divides each small screen by a serial number. Means, first coding means for performing high-efficiency coding by interframe coding for each of the small screens, the corresponding small screen number and the coded image signal are multiplexed and assembled into a packet, and an image is generated. First packet multiplexing means for transmitting to the image packet decoding device as a packet, and first packet decomposing means for receiving the missing number packet transmitted from the image packet decoding device and decomposing the received missing number packet , And means for forcibly intra-frame encoding a small screen corresponding to the missing number obtained by decomposition.

[0015]

When the screen dividing unit of the image packet encoding device divides the screen into small screens, a serial number is given to each small screen. Then, the receiving side monitors the continuity of the numbers and recognizes the numbers of the small screens discarded due to the interruption of the continuity.

When the image packet encoding device receives the missing number of the small screen from the first packet decomposing means,
The small screen corresponding to the missing number is forcibly intra-coded. That is, since the small screen of the succeeding screen corresponding to the missing number is processed by the intra-frame coding, the succeeding image can be continuously transmitted.

[0017]

Embodiments of the present invention will be described below with reference to the accompanying drawings. FIG. 1 is a configuration diagram of an image packet encoding device and a decoding device according to the present invention, FIG. 2 is a screen block diagram when an input image is divided into four, FIG. 3 is a diagram showing an image packet format, and FIG. 4 is missing. It is a figure which shows the format of a number packet.

First, the image packet coding apparatus will be described. The image packet encoding device assembles an image signal in which an input image is subjected to high efficiency encoding into a packet and transmits the image packet to a packet line.

The image packet encoding device 1 is configured to use the input image 1
00 is divided into a plurality of small screens (four small screens in the present embodiment), and a screen division circuit 2 for giving consecutive numbers 1 to 4 to each small screen.
And the divided small-screen image signal 101 with the screen number 1,
The encoding circuit 3 for performing high-efficiency encoding in the order of 2, 3, and 4, the small screen number signal 102, the encoded image signal 103 output from the encoding circuit 3, and the header including a predetermined destination address are packetized. A packet multiplexing circuit 4 which is assembled into the packet packet 200 and is transmitted to the packet line 200 in the format of the image packet 104 shown in FIG.
0 to the packet number decomposing circuit 5 for decomposing the missing number packet 105 transmitted in the format of the missing number packet 105 shown in FIG. 4 and inputting the missing number signal 106 to the encoding circuit 3. Further, the encoding circuit 3 further has a function of forcibly intra-frame encoding the small screen corresponding to the missing number when the missing number signal 106 is input from the packet disassembling circuit 5.

High-efficiency coding is performed by, for example, ITU-
Recommendation H.T. The information is compressed by performing motion compensation interframe coding or intraframe coding, DCT transform coding, quantization, or variable length coding adopted in H.261.

Next, the image packet decoding device will be described. The image packet decoding device 10 receives the image packet 104 transmitted from the image packet encoding device 1 via the packet line 200, and the image packet 104 is received.
4 coded image signal 201 and small screen number signal 20
A packet decomposing circuit 11 for decomposing into two, a decoding circuit 12 for decoding the coded image signal 201 and outputting a decoded image signal 203, and a receiving small screen number signal 202 from the packet decomposing circuit 11. The received small screen number signal 202 is output to the decoding circuit 12 as a small screen number signal 204 in the order of the received numbers, and it is detected that the continuity of the received numbers is interrupted, and the interrupted number is output as the missing number signal 20.
5, the small screen number detection circuit 13 that outputs 5 and the missing number signal 205 are assembled into a packet, and the missing number packet 1
05, and the packet multiplexing circuit 14 for transmitting to the image packet encoding device 1 via the packet line 206.

When the decoding circuit 12 recognizes the missing number from the small screen number 204 input from the small screen number detection circuit 13, the corresponding small screen area is frozen and the small screens other than the small screen are decoded. It further has a function of outputting as a decoded image signal 203.

Next, the operation of the image packet encoding device 1 and the decoding device 2 will be described.

First, the image 100 input to the image packet encoding device 1 is divided into four small screens by the screen division circuit 2, and each small screen is encoded by the encoding circuit 3. Next, each encoded small screen is assembled into a packet by the packet multiplexing circuit 4 together with the small screen number and the header, and the packet circuit 200 is formed as the image packet 104.
Be transmitted to.

On the other hand, the image packet 104 is input to the packet disassembling circuit 11 of the image packet decoding device 2,
It is decomposed into an image signal 201 and a small screen number signal 202. Then, the image signal 201 is output as a decoded image signal 203 via the decoding circuit 12.

The small screen number signal is input to the small screen number detection circuit 13. If no image packet is discarded on the packet line 200, the input small screen numbers are repeated in the order of 1, 2, 3, 4. here,
For example, when the image packet of small screen number 3 is discarded, the received small screen numbers are 1, 2, and 4, so the small screen number detection circuit 13 indicates that the image packet of small screen number 3 has been discarded. Can be detected. Then, the detected discarding small screen number is input to the decoding circuit 12 as the signal 204 and to the packet multiplexing circuit 14 as the signal 205.

Next, when the image packet of the small screen number 3 is discarded, the missing number packet 105 is input from the packet multiplexing circuit 14 to the packet disassembling circuit 5 of the image packet encoding device 1 via the packet line 206. It Then, the missing number signal 106 obtained by decomposing in the packet decomposing circuit 5 is input to the encoding circuit 3.

When the missing number signal 106 is input, the coding circuit 3 forcibly codes the image of the small screen number 3 to be transmitted next by the intraframe coding.

That is, when the encoding is performed by the inter-frame encoding, the transmission of the next image is interrupted by discarding the small screen number 3, but at least the image of the small screen 3 is intra-frame encoded. By encoding with, the next image can be transmitted immediately.

In the image packet decoding device 2, when the image of the small screen 3 is discarded and is in an error state, the decoding circuit 12 outputs an unidentifiable image. When the lack of the small screen 3 is recognized by the signal 204, the area of the missing small screen 3 is frozen in the decoding circuit 12 and the images are decoded only in the other small screen areas, so that the small screens 1 and 2 are decoded. , 4 can be output.

[0031]

According to the present invention, when a screen is divided into small screens by the screen dividing means, a continuous number is added to each small screen, and the receiving side monitors the continuity of this number and the continuity is interrupted. Since the number of the discarded small screen is recognized, the discarded small screen can be easily detected.

Further, the detected small screen number is returned to the transmitting side, and the transmitting side performs intra-frame coding by this information, whereby the reception error can be immediately recovered.

[Brief description of drawings]

FIG. 1 is a configuration diagram of an image packet encoding device and a decoding device according to the present invention.

FIG. 2 is a screen division diagram when an input image is divided into four.

FIG. 3 is a diagram showing a format of an image packet.

FIG. 4 is a diagram showing a format of a missing number packet.

[Explanation of symbols]

 1 Image Packet Coding Device 2 Screen Division Circuit 3, 12 Encoding Circuit 4, 14 Packet Multiplexing Circuit 5, 11 Packet Decomposing Circuit 10 Image Packet Decoding Device 13 Small Screen Number Detection Circuit 200, 206 Packet Line

Continuation of the front page (51) Int.Cl. ⁶ Identification code Office reference number FI technical display location H04N 7/15

Claims (4)

[Claims] 1. A screen dividing means for dividing an inputted image signal of one screen into a plurality of image signals of small screens and giving a serial number to each small screen, and high efficiency by inter-frame coding for each small screen. First packet multiplexing that encodes first encoding means, multiplexes the corresponding small screen number and the encoded image signal into a packet, and transmits it as an image packet to an image packet decoding device. Means and
First packet decomposing means for receiving a missing number packet transmitted from the image packet decoding device and decomposing the received missing number packet, and forcibly a small screen corresponding to the missing number obtained by decomposing. An image packet encoding device comprising: an intra-frame encoding unit. 2. The image packet coding apparatus according to claim 1, wherein the first coding means includes means for performing high-efficiency coding by intra-frame coding for each of the small screens. 3. A second packet decomposing means for receiving the image packet from the first packet multiplexing means, decomposing the received image packet, and separating the small image number and the encoded image signal. Decoding means for decoding the coded image signal, and missing number detecting means for detecting a missing number of the small screen based on the small screen number obtained by the second packet decomposing means. An image packet decoding device, comprising: a second packet multiplexing means for assembling the detected missing number into a packet and transmitting it to the image packet coding device as a missing number packet. 4. The decoding means, when receiving the notification of the missing number of the small screen from the missing number detecting means, freezes the missing small screen area and decodes the image only in other small screen areas. The image packet decoding device according to claim 3, wherein