• 首页
  • 期刊简介
  • 编委会
  • 投稿指南
  • 收录情况
  • 杂志订阅
  • 联系我们
引用本文:杜 梅,曹蔚然,赵怀慈.基于压缩感知的感兴趣区域编码[J].软件工程,2017,20(6):15-16.【点击复制】
【打印本页】   【下载PDF全文】   【查看/发表评论】  【下载PDF阅读器】  
←前一篇|后一篇→ 过刊浏览
分享到: 微信 更多
基于压缩感知的感兴趣区域编码
杜 梅,曹蔚然,赵怀慈1,2
1.(1.沈阳师范大学软件学院,辽宁 沈阳 110034;2.
2.中国科学院沈阳自动化研究所,辽宁 沈阳 110016)
摘 要: 受无线信道带宽的约束,图像经传统图像压缩方法压缩并经无线信道传输后,图像质量受损严重,会 影响后续探测识别结果的准确性,本文针对这个问题提出一种基于压缩感知(Compressed Sensing,CS)的感兴趣区域 (Region Of Interest,ROI)图像压缩方法。首先,将位平面位移技术引入压缩感知,对获得的图像进行量化、位平面分 解;然后位移ROI位平面,并给出编码方案;最后,在解码端通过解码、重构,得到ROI质量良好的重构图像。仿真 结果表明,本文算法重构图像的ROI部分PSNR高于传统的压缩感知编码方法,验证了方法的可行性和有效性,从而为 ROI图像编码提供了一种可行的解决方案。
关键词: 感兴趣区域;压缩感知;位平面位移;探测识别
中图分类号: TP751.1    文献标识码: A
基金项目: 沈阳师范大学科研基金(L201516);辽宁省自然科学基金项目(2013010420-401).
Region of Interest Coding Based on Compressed Sensing
DU Mei,CAO Weiran,ZHAO Huaici1,2
1.( 1.Software Institute, Shenyang Normal University, Shenyang 110034, China;2.
2.Shenyang Institute of Automation, Chinese Academy of Science, Shenyang 110016, China)
Abstract: In order to solve the conflict between the requirement for high image quality for detection and recognition task and the constraint of wireless channel bandwidth,the paper proposes a ROI(Region of Interest) coding algorithm based on compressed sensing.Compressed sensing is employed to compress images due to its excellent anti-interference capability,and bit plane scaling technology is introduced into it.First,the blocked compressed sensing signals are quantified and decomposed to bit planes.Then,the bit planes of the region of interest are shifted.Finally after decoding,the high-quality reconstructed ROI image is achieved.Experimental results indicate that the PSNR of the reconstructed ROI part encoded through the bit plane shifting algorithm is higher than that through the traditional compressed sensing coding technology. Therefore the ROI coding algorithm based on compressed sensing can improve the efficiency of the wireless image transmission system,and meet the requirements of target detection and identification better.
Keywords: ROI(Region Of Interest);compressed sensing;bit plane shifting;detection and identification


版权所有:软件工程杂志社
地址:辽宁省沈阳市浑南区新秀街2号 邮政编码:110179
电话:0411-84767887 传真:0411-84835089 Email:semagazine@neusoft.edu.cn
备案号:辽ICP备17007376号-1
技术支持:北京勤云科技发展有限公司

用微信扫一扫

用微信扫一扫