基于H.264視頻分區碼流聯(lián)合信源信道編碼技術(shù)研究
【論文學(xué)科】通信與信息系統論文
【論文級別】碩士論文,碩士畢業(yè)論文,碩士研究生論文
【中文題名】基于H.264視頻分區碼流聯(lián)合信源信道編碼技術(shù)研究
【英文題名】Joint Source Channel Coding Technique Research Based on H.264 Video Partitioning Code Stream
【所屬分類(lèi)】信息科技,電信技術(shù),通信
【中文關(guān)鍵詞】聯(lián)合信源信道編碼; 不等差錯保護; 差錯控制
【英文關(guān)鍵詞】Joint source channel coding; Unequal error protection; Error control
【中文摘要】本文主要研究了H.264編碼器生成的分區碼流數據的聯(lián)合信源信道編碼問(wèn)題,提出了針對高噪聲信道環(huán)境下的的聯(lián)合信源信道保護方法。本文首先對聯(lián)合信源信道編碼進(jìn)行了概述,分析了國內外關(guān)于聯(lián)合信源信道編碼的研究現狀。其次對視頻壓縮與差錯控制做了介紹,然后對H.264視頻編碼標準進(jìn)行了概述,并對其采用的各種關(guān)鍵技術(shù)進(jìn)行了深入的研究。在以上分析、研究的基礎之上,結合H.264中采用的數據分類(lèi)的抗誤碼技術(shù)設計了一種不等差錯保護方案。核心思想是在高噪聲信道環(huán)境下,通過(guò)對H.264編碼器生成的碼流進(jìn)行優(yōu)先級劃分,對重要數據加強信道保護,以獲得較好的視頻恢復效果。此外,在實(shí)際環(huán)境下,帶寬受限、高噪聲環(huán)境是視頻通信的難點(diǎn),本文進(jìn)一步提出一種方法,通過(guò)信道編碼和編碼端參數之間的有效匹配實(shí)現聯(lián)合信源信道保護的目的,取得了良好的誤碼保護的效果,用有限的冗余實(shí)現了有效的誤碼保護。最后,對全文的工作進(jìn)行了總結,并展望了今后(來(lái)源:Afa8a8aBC論文網(wǎng)www.abclunwen.com)有待進(jìn)一步完善的工作。(來(lái)源:A2323BC論文網(wǎng)www.abclunwen.com)
【英文摘要】 H.264 is a new video coding standard. It’s developed by united video group JVT which consists of Video Coding Experts Group(VCEG) of ITU-T and Movable Image Coding Experts Group(MPEG)of ISO/IEC. It uses many new coding techniques and can improve coding efficiency obviously.In recent years, with the rapid development of communication technology, as well as individual users on the new communications needs of the business growing, multimedia communications to be especially rapid development of video communications. However, due to tremendous volume of data and video information in the transmission must be carried out before compression to reduce the amount of information transmission channel. After the video compression coding information in the course of transmission is highly susceptible to channel conditions and other factors from the impact of error. Therefore, the adoption of effective anti-BER video coding technology to ensure reliable transmission of information is very necessary.Isolation Theory traditional assumptions source coding and channel coding complexity can be unlimited increase in coding design neglected some of the practical design of the design principle of the separation of the efficiency of the system may be lower, and this has stimulated a more effective joint source and channel coding methods of exploration and research. Joint Source-channel coding based on the state channel, weigh source coding and channel coding rate rate on the impact of distortion in the source and channel coding and distribution of redundancy, through a joint strategy in both coding appropriate allocation of resources, so as to achieve the best end-to-end results to achieve the best overall performance.This paper puts forward an unequal error protection plan. The strategy mainly researches on how to partition the information created from source coder, and uses the corresponding unequal error protection to make the important date bits be seized of more error renew ability, the aim is to obtain the best whole performance. The paper proposes that according to effect of the different parts of the code stream created from H.264 coder to the video, we can use the different error protection according to different bits. Those bits that are important to renew video use higher channel protection (use more intendance bits to protect information code stream), otherwise those bits that are not very important to renew video use lower channel protection (not very important means that if those bits are destroyed, it will not bring very severe influence or can renew those content with some compensate method). In DP mode, macro-block dates in a slice are separated into A、B、C three parts to transmit respectively. A date includes the most important information in decoding video; B date is more important than C dates; C date is the lowest important. Besides the three dates have these connections: B and C date are independent, but the decoding of those date must depend on A date. According to this partition mode, coder can use unequal error protection to different date selectively in order to guarantee the most important information can be received if the channel is not good. With the FEC protection towards H.264 code stream, we use FEC1 code that has the more ability to correct the error and redundancy to protect the most important dates, including SPS、PPS and A date, and use FEC2 code that has the less ability to correct the error and redundancy to protect the less important dates, including B and C date. With this unequal error protection, we guarantee the most important information correct renew under high error condition in order to obtain the better decoding effect.The experimental results show that the video transmission in the course of error ranging from protection mechanisms implemented, although an increase of redundancy, but it could be improved to a certain extent, the transmission system performance, improve video streaming channel in the worst anti-environment BER capacity. Stationary random error channel and restore the image PSNR-Y maximum value increased 8.73 dB at least 4.49 dB increase, objective quality improvement is obvious, and UEP Without UEP system than the performance gain access channel error probability with the increase increase. Gauss channel and restore the image PSNR-Y maximum value increased 9.36 dB, at least 2.71 dB increase. Restoration of images from the subjective feeling, the adoption of error ranging from the protection of the restoration of the subjective image quality has been greatly enhanced. The results can be seen through the introduction of the coding strategy UEP, although an increase of redundancy, but significantly improved the quality of the decoding of the video.After researches on unequal error protection scheme, the paper put forward a joint source channel coding strategy further aim at wireless channel’s characteristic that its bandwidth is limited and error is high. The thought is that channel coding must acclimatize itself to channel\'s state, matching H.264 source coder\'s QP、the packet\'s length to FEC code\'s rate. When the channel is not good, match coarse QP to less FEC code\'s rate. If increasing the value of QP, the code\'s rate is reduced, saving the source\'s bits; these bits can be used to reinforce the ability of FEC. The other way round, when the channel is good, match exquisite QP to higher FEC code\'s rate, used to transmit high quality video. Combining the match strategy with unequal error protection, we can achieve better balance with reliability and validity.The experiments results indicate that under high error environment using joint source channel coding can wipe off error effectively and also has better system performance than not using joint mechanism. Under high error environment, the PSNR Y at decoder receives improvement at high range. This can guarantee the video information reliable transmission and also can utilize the channel\'s limited bandwidth source effectively.In short, the joint source-channel video transmission technology in the protection of information, is an effective method, especially in bad channel conditions of decreased slowly, and with obvious superior system performance. Can guarantee reliable transmission of video information can be effective use of limited channel bandwidth resources. With limited redundancy to achieve an effective error protection, access to the consumption of redundancy and error a better balance between protections.
| 內容提要 | 4-7 |
| 第一章 緒論 | 7-15 |
| 1.1 論文選題背景 | 7-9 |
| 1.2 聯(lián)合信源信道編碼 | 9-13 |
| 1.2.1 聯(lián)合信源信道編碼概述 | 9-11 |
| 1.2.2 聯(lián)合信源信道編碼研究現狀 | 11-13 |
| 1.2.3 聯(lián)合信源信道編碼意義 | 13 |
| 1.3 課題研究的內容 | 13-15 |
| 第二章 視頻壓縮與差錯控制 | 15-22 |
| 2.1 視頻壓縮編碼概述 | 15-18 |
| 2.1.1 視頻序列冗余特性 | 15-16 |
| 2.1.2 視頻編碼技術(shù) | 16-18 |
| 2.2 差錯控制概述 | 18-22 |
| 2.2.1 誤碼產(chǎn)生原因及其影響 | 18 |
| 2.2.2 視頻通信差錯控制技術(shù) | 18-22 |
| 第三章 H.264 視頻編碼標準 | 22-32 |
| 3.1 H.264 編碼的主要目標 | 22 |
| 3.2 H.264 的基本框架 | 22-25 |
| 3.2.1 H.264 的檔次等級 | 22-23 |
| 3.2.2 H.264 編解碼器 | 23-25 |
| 3.3 H.264 的關(guān)鍵技術(shù) | 25-32 |
| 3.3.1 圖像分割 | 25-26 |
| 3.3.2 幀內預測 | 26-27 |
| 3.3.3 數據分類(lèi) | 27-28 |
| 3.3.4 參數設置 | 28-30 |
| 3.3.5 靈活的宏塊排序 | 30-32 |
| 第四章 H.264 視頻碼流的不等差錯保護方案 | 32-54 |
| 4.1 H.264 中句法元素組織結構 | 32-37 |
| 4.1.1 H.264 分層設計 | 32-33 |
| 4.1.2 Slice Layer設置 | 33-35 |
| 4.1.3 NAL設置 | 35-37 |
| 4.2 RTP打包規則 | 37-40 |
| 4.2.1 RTP包的組織結構 | 38-39 |
| 4.2.2 H.264 中RTP打包規則 | 39-40 |
| 4.3 不等差錯保護研究現狀 | 40 |
| 4.4 不等差錯保護方案設計與實(shí)現 | 40-46 |
| 4.4.1 聯(lián)合信源信道編碼原理 | 40-42 |
| 4.4.2 不等差錯保護方案算法描述 | 42-43 |
| 4.4.3 方案實(shí)現 | 43-46 |
| 4.5 實(shí)驗分析 | 46-54 |
| 4.5.1 實(shí)驗環(huán)境設置 | 46 |
| 4.5.2 算法性能測試與分析 | 46-54 |
| 第五章 H.264 視頻碼流的聯(lián)合信源信道編碼保護 | 54-70 |
| 5.1 碼率控制 | 54-59 |
| 5.1.1 碼率控制的重要性 | 54-55 |
| 5.1.2 碼率控制的基本原理 | 55 |
| 5.1.3 H.264 編碼參數對碼率的影響 | 55-59 |
| 5.2 聯(lián)合信源信道編碼設計與實(shí)現 | 59-61 |
| 5.2.1 聯(lián)合信源信道編碼算法描述 | 59-60 |
| 5.2.2 聯(lián)合信源信道編碼實(shí)現 | 60-61 |
| 5.3 實(shí)驗分析 | 61-70 |
| 5.3.1 實(shí)驗環(huán)境設置 | 61-63 |
| 5.3.2 算法性能測試與分析 | 63-70 |
| 第六章 總結與展望 | 70-72 |
| 6.1 全文總結 | 70-71 |
| 6.2 工作展望 | 71-72 |
| 參考文獻 | 72-76 |
| 摘要 | 76-79 |
| ABSTRACT | 79-81 |
| 致謝 | 82-83 |
| 導師及作者簡(jiǎn)介 | 83 |