阅读说明：本技术 一种帧级别的应用层动态fec编码方法 (frame-level application layer dynamic FEC encoding method ) 是由 陈平平 柳粟杰 宋金城 李然 王�锋 于 2019-10-18 设计创作，主要内容包括：本发明涉及一种帧级别的应用层动态FEC编码方法。利用视频传输系统,在发送端,视频序列先经过H.264视频编码器,设定量化参数(QP)值,视频编码器将每帧视频图像压缩成不同个数的源数据包,FEC编码将每帧的源数据包为一组,利用前向纠错编码算法生成一定个数的冗余包,二者组成一个FEC编码组。接收端可以根据对应的纠错译码算法恢复出源数据。本发明基于I帧和P帧在视频解码中具有相同的重要性,以及编码包个数多或码率低的FEC性能更好的原理,将包个数多的I帧码率提高,将其多出的冗余包分配给包个数少的P帧,不同包数的P帧有不同的码率,可使得I帧和P帧的译码效果相当,在解码中达到一致的重要性,提升视频的平均峰值信噪比(PSNR),使得视频整体质量上升。(The invention relates to a frame-level application layer dynamic FEC encoding method, which utilizes a video transmission system, a video sequence passes through an H.264 video encoder at a transmitting end, a Quantization Parameter (QP) value is set, the video encoder compresses each frame of video image into source data packets with different numbers, the FEC encoding divides the source data packets of each frame into groups, and a redundancy packet with a certain number is generated by utilizing a forward error correction encoding algorithm, and the two groups form FEC encoding groups.)

The method for dynamically FEC encoding of application layers at the level of frames is characterized in that firstly, at a sending end, each frame of video image in a video sequence is compressed into source data packets with different numbers through an H.264 video encoder, secondly, FEC encoding divides the source data packets of each frame of video image into groups in a frame-level encoding mode, meanwhile, redundancy packets corresponding to the source data packets of each frame of video image are generated through a forward error correction encoding algorithm, thirdly, the source data packets of each frame of video image and redundancy packets corresponding to the source data packets of each frame of video image are combined into FEC encoding groups, and finally, a dynamic FEC encoding mode related to the frame level is designed, so that the code rate of each frame is dynamically allocated according to the different numbers of the source data packets of each frame of video image, namely, the total number of the packets is guaranteed to be unchanged, the code rate of the I frame is improved, the number of the redundancy packets reduced by the I frame is allocated to the corresponding P frame, the code rate of the P frame is reduced, the decoding effect of the I frame and the average quality of the.

2. The frame-level application-layer dynamic FEC encoding method of claim 1, wherein at a receiving end, source data can be recovered according to a forward error correction decoding algorithm corresponding to a transmitting end.

Technical Field

The invention relates to channel coding in the communication field, in particular to an application layer dynamic FEC coding method of frame levels.

Background

In recent years, with the rapid development of wireless communication technology and internet technology, the demand for multimedia communication services has been increasing. With the construction of a fifth generation mobile communication (5G) network, the characteristics of high data transmission rate, low time delay, ultra-large network capacity and the like of the network meet the requirements of large data transmission, real-time application such as automatic driving and the like, and communication of the Internet of things.

The initial multimedia content is huge in data volume and limited in network bandwidth, and is usually downloaded to the local and then played. Later, due to the development of video coding technologies such as h.264, video was able to be streamed in real time. With the continuous development of the mobile internet era, people have increased the demand for wireless streaming media applications, and streaming media technology has the advantages of short start-up delay and the like, and is very suitable for the transmission of real-time multimedia applications, so the hot and difficult point of current research of researchers is how to ensure the service quality of streaming media in a complex network transmission environment, and improve the reliability and effectiveness of streaming media transmission without wasting network bandwidth.

The wireless video transmission technology is a challenging subject in the video transmission development technology, the wireless channel environment is very complex, noise in the channel can cause signal fading, and network packet loss can be caused, so that a receiving end can generate error decoding during video decoding, and the video quality reconstructed by a video decoder is greatly reduced.

The basic principle of the FEC is that a transmitting end generates quantitative redundant packets according to an error correction coding algorithm, so that the error correction capability of is achieved, and the receiving end can restore lost data packets according to an error correction decoding algorithm.

FEC is techniques for recovering data lost in a network transmission process by adding a redundant part to transmitted data information based on an algorithm, and with the adoption of error correction coding algorithms such as fountain codes, RS codes and LDPC codes, engineering experiments prove that the technical coding has excellent performance, and FEC does not need a feedback channel, so that the FEC has the characteristics of low time delay and high reliability.

Disclosure of Invention

The invention aims to provide an application layer dynamic FEC encoding method of frame levels, which is used for independently encoding each frame and adopts different code rates, so that I frames and P frames with similar importance to visual reconstruction have close error correction performance, and the video is better reconstructed, and the overall quality of the video is better.

The technical scheme includes that frame-level application layer dynamic FEC encoding methods include that firstly, at a sending end, each frame of video image in a video sequence is compressed into source data packets with different numbers through an H.264 video encoder, secondly, FEC encoding divides the source data packets of each frame of video image into groups in a frame-level encoding mode, meanwhile, redundancy packets corresponding to the source data packets of each frame of video image are generated through a forward error correction encoding algorithm, thirdly, the source data packets of each frame of video image and the redundancy packets corresponding to the source data packets of each frame of video image are combined into FEC encoding groups, and finally, a frame-level dynamic FEC encoding mode is designed, so that the code rate of each frame is dynamically allocated according to different numbers of the source data packets of each frame of video image, namely, the total number of the packets is guaranteed to be unchanged, the code rate of an I frame is increased, the reduced redundancy packet number decreased by the I frame is allocated to a corresponding P frame, the code rate of the P frame is reduced, the decoding effects of the I frame and the P frame are equivalent, and the average quality of video transmission and the whole.

In the embodiment of the present invention, at the receiving end, the source data can be recovered according to the fec decoding algorithm corresponding to the transmitting end.

Compared with the prior art, the method has the advantages that dynamic FEC encoding schemes which encode according to frames are adopted to independently encode each frame, and different code rates are adopted, so that I frames and P frames which have similar importance on visual reconstruction have close error correction performance, the video is reconstructed better, and the overall quality of the video is better.

Drawings

FIG. 1 is a schematic diagram of a system model.

Fig. 2 is a schematic diagram of the FEC channel coding principle.

Fig. 3 is a diagram comparing video frames of a conventional scheme and a dynamic FEC scheme.

Detailed Description

The technical scheme of the invention is specifically explained below with reference to the accompanying drawings. The present embodiment is implemented on the premise of the technical solution of the present invention, and a detailed implementation manner and a specific operation process are given, but the scope of the present invention is not limited to the following embodiments.

The invention provides an frame-level application layer dynamic FEC encoding method, which comprises the steps of firstly compressing each frame of video image in a video sequence into source data packets with different numbers by an H.264 video encoder at a sending end, secondly dividing the source data packets of each frame of video image into groups by the FEC encoding method, simultaneously generating redundant packets corresponding to the source data packets of each frame of video image by using a forward error correction encoding algorithm, then forming FEC encoding groups by combining the source data packets of each frame of video image and the redundant packets corresponding to the source data packets of each frame of video image, and finally designing a dynamic FEC encoding method related to the frame level to ensure that the code rate of each frame is dynamically distributed according to the different numbers of the source data packets of each frame of video image, namely ensuring the total number of the packets to be unchanged, improving the code rate of the I frame, distributing the reduced redundant packets of the I frame to the corresponding P frame, reducing the code rate of the P frame, ensuring that the decoding effect of the I frame and the P frame is equivalent, and the average quality of video transmission is improved.

The following is a specific implementation of the present invention.

As shown in fig. 1, the figure shows the entire video transmission system. The method comprises the steps that a video source sequence is input at a sending end, compressed into a video stream through a video encoder, and then added with redundant data packets through an FEC channel encoder, a source data packet and the redundant data packets jointly form an FEC encoding set and transmit the data packets through a channel, the source data packet is recovered through an FEC channel decoder at a receiving end, and the video sequence is reconstructed through the video decoder.

The source video sequence is compressed into a video stream by an H.264 video encoder, the whole video sequence is divided into a plurality of GOPs, each GOP is provided with 1I frame and a plurality of P frames, the K source data packets of each frame are subjected to fixed code rate by adopting a frame-level channel coding mode, and (n-k) redundant packets are generated by utilizing an LDPC-steady application layer forward error correction coding algorithm to jointly form an (n, k) FEC coding group.

The specific allocation method comprises the steps of setting GOP size and QP size of an H.264 video encoder, improving the code rate of the I frame of the original traditional scheme, reducing the number of I frame redundant data packets compared with the traditional scheme, saving partial redundant data packets, allocating the redundant data packets saved by the I frame to the P frame with less source packets in a large proportion and allocating the redundant data packets saved by the I frame to the P frame with more source packets in a small proportion on the basis of the traditional scheme, so that the P frames with different source data packets also determine different code rates, and different frames have different coding rates.

The transmission system adopts an H.264 video coding standard, the size of a GOP is set to be 25, namely I frames are inserted into every 25 frames, the size of a QP is set to be 30, the frame rate of a video sequence is 30 frames/second, and a packet loss model of the simulation is a random packet loss model, the video sequence used by the simulation is ParkScene, and two experimental schemes adopted by the simulation are a coding mode of frame coding and are respectively a traditional FEC coding scheme and dynamic FEC coding, the traditional FEC coding scheme is that the FEC code rate of each frame is equal no matter whether the I frame or the P frame, the FEC code rate of the simulation is set to be 1/2, the dynamic FEC coding scheme is that on the basis of the traditional FEC coding scheme, the total number of the packets is ensured to be unchanged, the code rate of the I frame is improved, the redundant data packets are allocated to the P frame, the code rate of the P frame is reduced, k is the number of source data packets of the ParkScene sequence after the H.264 video coding, and the dynamic FEC coding scheme of the simulation at the frame level:

1) when k is more than 100, the frame is an I frame, and the FEC encoding code rate is 2/3;

2) when 80< = k <90, the frame is a P frame, and the FEC coding rate is 1/3;

3) when 70< = k <80, the frame is a P frame, and the FEC coding rate is 18/25;

4) when 60< = k <70, the frame is a P frame, and the FEC coding rate is 3/5;

5) when 50< = k <60, the frame is a P frame, and the FEC coding rate is 1/6;

6) when 40< = k <50, the frame is a P frame, and the FEC coding rate is 1/6.

The transmission system performs transmission verification with PLR (packet loss rate) of 1%, 5%, and 10% using two experimental schemes, respectively, and the experimental results are shown in table 1 below:

as can be seen from table 1, the error-resilient performance of the dynamic FEC coding scheme according to the frame level of the present invention is significantly better than that of the conventional FEC coding scheme with the same code rate for each frame of video, and the PSNR values of the method of the present invention for the ParkScene video sequence are respectively improved by 0.3414dB, 1.1074dB and 0.994dB compared with that of the conventional FEC coding scheme under the condition that the packet loss rates are respectively 1%, 5% and 10%. Therefore, the invention effectively improves the reconstructed video quality of the FEC video transmission system.

As shown in fig. 3, an FEC scheme is adopted with a packet loss rate of 5%, and a certain frame picture of a decoded video frame received at a receiving end is a received frame picture of a conventional scheme above the received frame picture and a received frame picture of the scheme of the present patent below the received frame picture.

The above are preferred embodiments of the present invention, and all changes made according to the technical scheme of the present invention that produce functional effects do not exceed the scope of the technical scheme of the present invention belong to the protection scope of the present invention.