基于Bandelet稀疏的移动机器人环境视觉纹理图像的压缩传感与重构

摘要/Abstract

摘要：

压缩传感技术为移动机器人环境视觉的实时高效处理与传输提供了一种新的解决方法。结合Bandelet变换自适应跟踪图像正则方向的特点,进行了基于Bandelet稀疏和正交匹配追踪(OMP)算法的环境纹理图像压缩传感重构分析研究。结果表明：在较大观测值下,Bandelet稀疏重构与传统小波稀疏重构效果差别不大;在较小观测值下,传统sym8小波稀疏重构出现不稳定状态,出现块状信息缺失,不能有效重构,而Bandelet稀疏重构效果相对稳定;在给定观测值下,Bandelet稀疏重构的边缘细节表达能力优于sym8小波,说明Bandelet变换在压缩传感采样高压缩比下恢复重构具有有效性和稳定性。

关键词: 机器人环境视觉;图像重构;压缩传感, Bandelet稀疏

Abstract:

The occurrence of compressed sampling provided a novel efficient, real-time processing methodology for robot environmental vision mass informations. Combined with the Bandelet transform's advantage of adaptive tracking regularization direction of image, the research of the compressed sampling and reconstruction of environmental texture images was carried out herein, which was based on the Bandelet transform and orthogonal matching pursuit (OMP) algorithm. The experimental results show: the reconstruction results with Bandelet sparsity (BS) and wavelet sparsity (WS) are similar with a bigger measurement value; however, with a smaller measurement value, the reconstruction based on WS takes on an unstable state and sometimes the image can not be reconstructed effectively; with the same given measurement value, the reconstruction ability of edge details in the image based on BS is better than that of WS. The results illustrate that under higher compressive sensing compression ratio, the BS based environmental texture image reconstruction has a better effectiveness and stability than that of WS based reconstruction.

Key words: robot environmental vision, image reconstruction, compressive sensing, Bandelet sparsity(BS)

中图分类号:

马如远, 刘继忠, 金明亮, 柴国钟, 王光辉. 基于Bandelet稀疏的移动机器人环境视觉纹理图像的压缩传感与重构[J]. 中国机械工程, 2014, 25(11): 1535-1540.

Ma Ruyuan, Liu Jizhong, Jin Mingliang, Chai Guozhong, Wang Guanghui. Bandelet Sparsity Based Compressive Sensing and Reconstruction of Texture Images for Mobile Robot Environmental Visions[J]. China Mechanical Engineering, 2014, 25(11): 1535-1540.

参考文献

[1]马如远，金明亮，刘继忠，等. 移动机器人环境视觉小波稀疏压缩传感和识别[J]. 传感技术学报，2012，25（4）：519-523.
Ma Ruyuan, Jin Mingliang, Liu Jizhong, et al. Wavelet Sparsity Based Compressive Sensing and Direct Recognition for Mobile Robot Environmental Vision[J]. Chinese Journal of Sensors and Actuators, 2012, 25(4): 519-523.
[2]赵士彬，姚素英，徐江涛. 基于压缩感知的低功耗高效率CMOS图像传感器设计[J]. 传感技术学报，2011，24（8）：1151-1157.
Zhao Shibin, Yao Suying, Xu Jiangtao. Low Power High CMOS Image Sensor Design Based on Compressed Sensing[J]. Chinese Journal of Sensors and Actuators, 2011, 24(8): 1151-1157.
[3]Wang Y, Bermak A, Boussaid F. FPGA Implementation of Compressive Sampling for Sensor Network Applications[C]//Proceedings of the 2nd Asia Symposium on Quality Electronic Design. Penang, Malaysia, 2010: 5-8.
[4]王向阳，杨红颖. 一种基于小波包变换的纹理图像压缩算法[J]. 测绘学报，2004，33（3）：239-243.
Wang Xiangyang, Yang Hongying. A New Wavelet Packet Coding Algorithm for Texture-rich Images[J]. Acta Geodaetica et Cartographica Sinica, 2004, 33(3): 239-243.
[5]韩培友，张雯，郝重阳，等. 一种基于混合编码的图像纹理压缩方法[J]. 微电子学与计算机，2006，23 (11)：178-180，184.
Han Peiyou, Zhang Wen, Hao Chongyang, et al. A Texture Compression Algorithm Based on Hybrid Coding[J]. Microelectronics & Computer, 2006, 23 (11): 178-180, 184.
[6]王超，叶凤琴，叶中付. 基于结构纹理分解的改进图像压缩算法[J]. 中国科学技术大学学报，2007，37（12）：1449-1454.
Wang Chao, Ye Fengqin, Ye Zhongfu. An Improved Image Compression Method Using Cartoon-texture Decomposition[J]. Journal of University of Science and Technology of China, 2007, 37(12): 1449-1454.
[7]潘志刚，高鑫. 针对纹理图像压缩的改进SPIHT算法[J]. 中国科学院研究生院学报，2010，27（2）：222-227.
Pan Zhigang, Gao Xin. An Improved SPIHT Algorithm for Texture Image Compression[J]. Journal of Graduate School of the Chinese Academy of Sciences, 2010, 27(2): 222-227.
[8]张军，成礼智，杨海滨，等. 基于纹理的自适应提升小波变换图像压缩[J]. 计算机学报，2010，33（1）：184-192.
Zhang Jun, Cheng Lizhi, Yang Haibin, et al. Adaptive Lifting Wavelet Transform and Image Compression via Texture[J]. Chinese Journal of Computers, 2010, 33(1): 184-192.
[9]Hari Hara Santosh D, Dasari P, Kiran N L S, et al. FRCT Based Efficient Image Compression for Texture Images[C]//2012 International Conference on Computing, Communication and Applications. Dindigul, 2012: 1-6.
[10]朱梅，李章维. 基于Bandelets域的自适应图像压缩[J]. 计算机工程，2011，37（7）：241-242，252.
Zhu Mei, Li Zhangwei. Adaptive Image Compression Based on Bandelets Domaiin[J]. Computer Engineering, 2011, 37(7): 241-242, 252.
[11]田润澜，肖卫华，齐兴龙. 几种图像变换算法性能比较[J]. 吉林大学大学报：信息科学版，2010，28（5）：439-444.
Tian Runlan, Xiao Weihua, Qi Xinglong. Comparion of Several Image Transform[J]. Journal of Jilin University (Information Science Edition), 2010, 28(5): 439-444.
[12]金坚，谷源涛，梅顺良. 压缩采样技术及其应用[J]. 电子与信息学报，2010，32(2)：470-475.
Jin Jan, Gu Yuantao, Mei Shunliang. An Introduction to Compressive Sampling and Its Applications[J]. Journal of Electronics & Information Technology, 2010, 32(2): 470-475.
[13]Iqbal M, Chen J. Compressive Sampling of Natural Images Using Bandelet Basis[C]//2011 4th International Conference on Image and Signal Processing. Shanghai, 2011: 958-962.
[14]Donoho D L,Huo X M.Uncertainty Principles and Ideal Atomic Decomposition[J]. IEEE Transactions on Information Theory, 2001, 47: 2845-2862.
[15]Donoho D L, Tsaig Y. Extensions of Compressed Sensing[J]. Signal Processing, 2006, 47: 549-571.