移动机器人环境视觉图像小波稀疏压缩传感与恢复重构.pdfVIP

Vo1.35 No.3 南昌大学学报(工科版) 第 35 卷第3 期 Journal of Nanchang University( Engineering Technology) Sept.2013 2013 年9 月 文章编号:l(脱-倒56(2013 )03 -0267 -04 移动机器人环境视觉图像小波稀疏 压缩传感与恢复重构 刘继忠1 ，金明亮1 ，马如远2 ，柴国钟2 ，王光辉1 ， 3 (1.南昌大学机电工程学院，江西南昌 33ω31 ;2. 浙江工业大学机械学院，浙江杭州 31饭削; 3. 滑铁卢大学系统设计工程系，滑铁卢 N2L3日) 摘要:基于最新压缩传感技术理论，采用小波稀疏方法和正交匹配跟踪算法，对移动机器人常规环境视觉图像 压缩传感存储与恢复重构进行了研究。结果表明:进行小波稀疏的压缩传感编码可以大大减少机器人环境视觉图 像信息量，降低存储与传输代价，通过正交匹配追踪算法恢复的视觉图像，可以满足机器人常规环境视觉探测。 关键词:机器人;环境视觉;小波稀疏;图像重构;压缩传感 申圄分类号:TP274; TP242.6;1凹01 文献标志码:A Wavelet Sparsity ßased Compressive Sensing and Reconstruction for Mobile Robot Environmental Vision 1 1 2 1 UU Ji-zhong ,JIN Ming-liang ,MA Ru-yuan ,CAI Guo-zhong2 ,WANG Guang-hui ,3 (1. Sch∞1 of Mechanical and Electrical Engineering ,Nanchang UIÚversity ,Nanchang 33∞31 ， China; 2. School of Mecharúcal Engineering ,Zhejiang UIÚversity of Technology ,Hangzhou 3川剧，China; 3. Depar恤ent of SYDE ,UIÚversity of Waterl∞，Waterloo N2口Gl ， Canada) Abstract: ßased on the novel compressed sampling theo巧， compressed sensing sωrage and reconstruction for environment images from robot vision were researched in the paper by the help of wavelet sp缸sity and orthogonal matching pursuit algorithm. Experiments of several typical environment images for sparsity ,compressed storage ,and reconstruction were c缸甘ed out. 咀le result showed it could reduce the amount of environment images information , and lower 也e cost of information storage and transmission. The reconstruction im