速记式加密算法在裸眼3D图像中的应用研究
发布时间：2018-08-17 16:42
【摘要】：随着计算机科学技术的不断发展,尤其是视频编解码技术的不断进步,立体视频的应用场景越来越广泛,加上人们对3D视频也越来越感兴趣,立体视频的编解技术得到空前发展。但是现阶段得到广泛使用的立体技术依然存在着缺陷,例如,在播放立体视频的时候,人们需要佩戴辅助的特制眼镜才能观赏到立体效果,并且,佩戴观影太久还会导致头晕目眩等身体不适的状态,严重的阻碍了立体视频的发展,也就导致了裸眼3D视频的萌芽。同样,对于裸眼3D技术来说,它本身的一些技术难点还没有被攻破,比如网络大数据的传输问题、海量数据的编解码融合问题、裸眼3D视频不能被直接编码存储的问题,甚至是硬件方向所面临的问题等。虽然现阶段已有以MVC视频编码技术为代表的很多优秀多视点编码技术,但MVC技术在编码多个视频时,需要把多个视频文件编码为单独的视频文件,因此裸眼3D中的大数据传输问题也并没有得到解决。本文主要研究如何解决大数据在网络中的传输问题,根据本文的方法可以实现在网络上不利用任何的附加资源使得裸眼3D可以以传统的2D视频的方式传输数据信息。论文主要是以H.264为视频编解码基础,利用速记式加密算法把深度图的数据信息隐藏到视频数据中,基于这样的思想实现了一个完整的裸眼3D视频和深度信息同时编解码的系统。本文主要涉及的工作如下:本文首先从整体框架上讲解了视频编解码设计过程,及详细介绍了速记式加密算法的来历及思想,并从H.264的编解码结构上剖析使用速记式加密算法的可能性。然后则详细的解释了LSB算法的原理,并且实现了以X.264中添加隐藏深度图像数据的模块,然后利用FFMPEG中的264解码库添加了提取深度信息的模块,并且从理论上分析出改进型的LSB算法在本文中的利用有两大主要缺点------载体容量不够和一帧深度图对应多帧视频。因此引用和改进了BPCS算法,根据BPCS算法在X.264和FFMPEG中实现了对深度信息的隐藏和提取处理,最后得出结果,并与LSB算法做了详细的对比,表明BPCS算法解决了LSB算法的两大缺点,同时还提高了融合之后裸眼3D图片的质量效果。最后还利用视频图像的客观评价方式通过数学模型计算表明利用BPCS算法实现的深度图像隐藏技术表现优秀。
[Abstract]:With the continuous development of computer science and technology, especially the progress of video coding and decoding technology, stereo video applications are becoming more and more widespread, coupled with the growing interest in 3D video, stereo video coding and decoding technology has been unprecedented development. When playing stereoscopic video, people need to wear special glasses to see the stereoscopic effect, and wearing too long can also lead to dizziness and other physical discomfort, seriously hindering the development of stereoscopic video, which led to the germination of naked-eye 3D video. Similarly, for naked-eye 3D technology, it itself. Some technical difficulties have not been overcome, such as the transmission of large network data, the fusion of massive data encoding and decoding, the problem that naked-eye 3D video can not be directly encoded and stored, and even the problem facing the hardware direction. MVC technology needs to encode multiple video files into a single video file when encoding multiple videos, so the problem of large data transmission in naked-eye 3D has not been solved. In this paper, H.264 is used as the basis of video coding and decoding, and shorthand encryption algorithm is used to hide the depth map data into the video data. Based on this idea, a complete system of coding and decoding the depth information and the naked-eye 3D video simultaneously is realized. The main work involved in this paper is as follows: Firstly, the design process of video encoding and decoding is explained from the overall framework, and the origin and idea of shorthand encryption algorithm is introduced in detail, and the possibility of using shorthand encryption algorithm is analyzed from the encoding and decoding structure of H.264. The module of hiding depth image data is added in X.264, and then the module of extracting depth information is added in the 264 decoding library of FFMPEG. The two main shortcomings of the improved LSB algorithm in this paper are analyzed theoretically, that is, insufficient carrier capacity and one depth map correspond to multi-frame video. BPCS algorithm, according to BPCS algorithm in X.264 and FFMPEG to achieve the hiding and extraction of depth information processing, and finally get the results, and compared with the LSB algorithm in detail, shows that BPCS algorithm solves the two major shortcomings of LSB algorithm, but also improves the quality of 3D images after fusion. The evaluation method shows that the depth image hiding technology based on BPCS algorithm performs well through mathematical model calculation.
【学位授予单位】：电子科技大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TN919.81
本文编号：2188268
[Abstract]:With the continuous development of computer science and technology, especially the progress of video coding and decoding technology, stereo video applications are becoming more and more widespread, coupled with the growing interest in 3D video, stereo video coding and decoding technology has been unprecedented development. When playing stereoscopic video, people need to wear special glasses to see the stereoscopic effect, and wearing too long can also lead to dizziness and other physical discomfort, seriously hindering the development of stereoscopic video, which led to the germination of naked-eye 3D video. Similarly, for naked-eye 3D technology, it itself. Some technical difficulties have not been overcome, such as the transmission of large network data, the fusion of massive data encoding and decoding, the problem that naked-eye 3D video can not be directly encoded and stored, and even the problem facing the hardware direction. MVC technology needs to encode multiple video files into a single video file when encoding multiple videos, so the problem of large data transmission in naked-eye 3D has not been solved. In this paper, H.264 is used as the basis of video coding and decoding, and shorthand encryption algorithm is used to hide the depth map data into the video data. Based on this idea, a complete system of coding and decoding the depth information and the naked-eye 3D video simultaneously is realized. The main work involved in this paper is as follows: Firstly, the design process of video encoding and decoding is explained from the overall framework, and the origin and idea of shorthand encryption algorithm is introduced in detail, and the possibility of using shorthand encryption algorithm is analyzed from the encoding and decoding structure of H.264. The module of hiding depth image data is added in X.264, and then the module of extracting depth information is added in the 264 decoding library of FFMPEG. The two main shortcomings of the improved LSB algorithm in this paper are analyzed theoretically, that is, insufficient carrier capacity and one depth map correspond to multi-frame video. BPCS algorithm, according to BPCS algorithm in X.264 and FFMPEG to achieve the hiding and extraction of depth information processing, and finally get the results, and compared with the LSB algorithm in detail, shows that BPCS algorithm solves the two major shortcomings of LSB algorithm, but also improves the quality of 3D images after fusion. The evaluation method shows that the depth image hiding technology based on BPCS algorithm performs well through mathematical model calculation.
【学位授予单位】：电子科技大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TN919.81
【参考文献】
相关期刊论文 前1条
1 王蕾睿;沈礼权;张兆扬;;一种基于AVS视频编码的快速帧间预测模式选择算法[J];中国图象图形学报;2007年10期
相关硕士学位论文 前1条
1 于凤利;2D-3D视频转换中深度图生成方法研究[D];山东大学;2012年
，本文编号：2188268
本文链接：https://www.wllwen.com/kejilunwen/wltx/2188268.html
