基于压缩感知的图像自适应编码及重构方法研究
发布时间：2019-04-18 18:39
【摘要】：随着数字媒体采集、显示以及处理技术的飞速发展,各种高质量图像及视频的新应用和服务不断出现,使得图像/视频数据呈爆炸式增长。海量的图像/视频数据给传输和存储提出了极高要求,如何实现高效压缩已成为图像及视频编解码领域长期存在的挑战性问题。近年来,新兴的压缩感知理论大大提高了信号的压缩率,降低了信号存储和传输的压力,这对于图像及视频编解码领域的研究无疑是一个大的革新和进步。本文简要介绍了现有的以压缩感知理论为基础的图像编码算法,重点介绍了分块压缩感知理论、稀疏性判定准则以及利用图像空间相关性的自适应图像编码算法。在此基础上,本文提出了一种图像自适应编码算法和两种针对图像序列重构的改进算法,具体内容为:(1)基于压缩感知的图像自适应编码算法:本文在满足编码端采样率要求的情况下,根据图像块在TV域的稀疏性,为每一个图像块合理地分配不同的采样率,以此提高图像的压缩率,同时解码端也能得到高质量的重构图像。(2)基于自适应卡尔曼的时域增强算法:本文在分块视频压缩感知MC-BCS-SPL算法的基础上,分析了图像序列中每幅图像的噪声分布特征,将自适应卡尔曼滤波思想运用到图像序列时域增强中,对重构后的图像在时域方向进行滤波,有效去除了帧间噪声,使图像的主观效果有了一定提高。(3)基于TVAL3的图像序列冗余重构算法:本文提出一种将TVAL3及新三步搜索法相结合的重构算法来实现图像序列的重构。在该算法中将TVAL3算法作为图像重建的算法,采用新三步搜索法(NTSS)作为块匹配算法,来得到当前帧在参考帧中的最优匹配块。利用上述方法对图像序列进行重构后,需要对其再进行维纳滤波,来得到更好的主观图像。实验结果表明本文提出的基于压缩感知的自适应编码算法有效减少了编码端的采样数据,实现了高效压缩;同时两种针对图像序列的改进重构算法也有效提高了图像序列的重构质量。
[Abstract]:With the rapid development of digital media acquisition, display and processing technology, a variety of high-quality image and video applications and services continue to appear, resulting in an explosive growth of image / video data. The huge amount of image / video data requires the transmission and storage. How to achieve efficient compression has become a long-standing challenge in the field of image and video coding and decoding. In recent years, the emerging compression sensing theory has greatly improved the compression ratio of signals and reduced the pressure of signal storage and transmission, which is undoubtedly a great innovation and progress in the field of image and video coding and decoding. In this paper, the existing image coding algorithms based on the compression perception theory are briefly introduced, and the block compression perception theory, the sparsity criterion and the adaptive image coding algorithm based on the spatial correlation of the image are emphatically introduced. On this basis, this paper proposes an adaptive image coding algorithm and two improved algorithms for image sequence reconstruction. The main contents are as follows: (1) Image adaptive coding algorithm based on compression perception: in this paper, according to the sparseness of image blocks in TV domain, different sampling rates are reasonably allocated to each image block under the condition that the sampling rate at the coding end is satisfied. In order to improve the compression ratio of the image, high quality reconstructed image can be obtained at the same time. (2) time domain enhancement algorithm based on adaptive Kalman: in this paper, based on the block video compression sensing MC-BCS-SPL algorithm, The noise distribution characteristics of each image in the image sequence are analyzed. The adaptive Kalman filter is applied to the time domain enhancement of the image sequence. The reconstructed image is filtered in the time domain direction, and the inter-frame noise is effectively removed. The subjective effect of image is improved. (3) redundant reconstruction algorithm of image sequence based on TVAL3: in this paper, a reconstruction algorithm combining TVAL3 and new three-step search method is proposed to realize the reconstruction of image sequence. In this algorithm, the TVAL3 algorithm is used as the image reconstruction algorithm, and the new three-step search method (NTSS) is used as the block matching algorithm to obtain the optimal matching block of the current frame in the reference frame. In order to obtain better subjective image, Wiener filtering is needed to reconstruct the image sequence using the above methods. The experimental results show that the proposed adaptive coding algorithm based on compression sensing can effectively reduce the sampling data at the coding end and achieve efficient compression. At the same time, two improved reconstruction algorithms for image sequences also effectively improve the reconstruction quality of image sequences.
【学位授予单位】：天津大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TP391.41
[Abstract]:With the rapid development of digital media acquisition, display and processing technology, a variety of high-quality image and video applications and services continue to appear, resulting in an explosive growth of image / video data. The huge amount of image / video data requires the transmission and storage. How to achieve efficient compression has become a long-standing challenge in the field of image and video coding and decoding. In recent years, the emerging compression sensing theory has greatly improved the compression ratio of signals and reduced the pressure of signal storage and transmission, which is undoubtedly a great innovation and progress in the field of image and video coding and decoding. In this paper, the existing image coding algorithms based on the compression perception theory are briefly introduced, and the block compression perception theory, the sparsity criterion and the adaptive image coding algorithm based on the spatial correlation of the image are emphatically introduced. On this basis, this paper proposes an adaptive image coding algorithm and two improved algorithms for image sequence reconstruction. The main contents are as follows: (1) Image adaptive coding algorithm based on compression perception: in this paper, according to the sparseness of image blocks in TV domain, different sampling rates are reasonably allocated to each image block under the condition that the sampling rate at the coding end is satisfied. In order to improve the compression ratio of the image, high quality reconstructed image can be obtained at the same time. (2) time domain enhancement algorithm based on adaptive Kalman: in this paper, based on the block video compression sensing MC-BCS-SPL algorithm, The noise distribution characteristics of each image in the image sequence are analyzed. The adaptive Kalman filter is applied to the time domain enhancement of the image sequence. The reconstructed image is filtered in the time domain direction, and the inter-frame noise is effectively removed. The subjective effect of image is improved. (3) redundant reconstruction algorithm of image sequence based on TVAL3: in this paper, a reconstruction algorithm combining TVAL3 and new three-step search method is proposed to realize the reconstruction of image sequence. In this algorithm, the TVAL3 algorithm is used as the image reconstruction algorithm, and the new three-step search method (NTSS) is used as the block matching algorithm to obtain the optimal matching block of the current frame in the reference frame. In order to obtain better subjective image, Wiener filtering is needed to reconstruct the image sequence using the above methods. The experimental results show that the proposed adaptive coding algorithm based on compression sensing can effectively reduce the sampling data at the coding end and achieve efficient compression. At the same time, two improved reconstruction algorithms for image sequences also effectively improve the reconstruction quality of image sequences.
【学位授予单位】：天津大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TP391.41
【相似文献】
相关期刊论文 前10条
1 孙英皓;唐棣;;一种有效的交互式图像局部内容放大方法[J];计算机工程与应用;2012年12期
2 李瑞帅;余淑华;潘凯;王圆;;图像特征金字塔快速计算方法[J];电子世界;2014年07期
3 曾喜良;王金娟;;指纹分割的块图像梯度因子聚类法[J];计算机与数字工程;2008年07期
4 刘陈;王欣欣;李凤霞;赵相坤;;一种快速保边的图像对象分割方法[J];北京理工大学学报;2010年02期
5 朱薇;刘利刚;;图像适应算法中非冗余显著图的计算[J];中国图象图形学报;2011年08期
6 柳有权;吴宗胜;韩红雷;吴恩华;;线条增强的建筑物图像抽象画生成[J];计算机辅助设计与图形学学报;2013年09期
7 吴骏,唐红梅,肖志涛,贾志成;一种基于相位信息的图像对称性检测方法[J];信号处理;2004年01期
8 邵静;高隽;赵莹;张旭东;;一种基于图像固有维度的感知物体检测方法[J];仪器仪表学报;2008年04期
9 潘如如;高卫东;;高紧度机织物图像倾斜的自动纠正[J];纺织学报;2009年10期
10 刘贵喜,赵曙光,杨万海;基于梯度塔形分解的多传感器图像融合[J];光电子·激光;2001年03期
相关会议论文 前10条
1 张一鸣;刘亚t，
本文编号：2460249
本文链接：https://www.wllwen.com/kejilunwen/ruanjiangongchenglunwen/2460249.html
