基于感兴趣区域的红外目标图像压缩研究  

Research on Infrared Target Image Compression Based on Region of Interest

作　　者：袁胜智[1] 谢晓方[1] 郭清风[1] 王建国[1] 朱锐[1] 

机构地区：[1]海军航空工程学院兵器科学与技术系,山东烟台264001

出　　处：《激光与红外》2008年第8期822-825,829,共5页Laser & Infrared

基　　金：海军预先研究课题基金

摘　　要：为了解决低带宽信道传输红外目标图像的问题,提出了一种逐步逼近二维最大熵阈值递推搜索算法,并将其应用到红外目标图像感兴趣区域的自动提取过程中,还以该算法为基础探讨了一种基于自动提取感兴趣区域的红外图像压缩方法,在JPEG2000框架下进行了具体实现,通过图像试验充分验证了该方法的有效性、实时性,具有重要的应用价值。In order to solve the transmission question for the infrared images through the low bandwidth data link,a new stepwise approaching and recurring threshold search algorithm based on 2D （two-dimension） maximum entropy principle was proposed for ROI automatically detected while a compressed scheme based on ROI was studied. An ex- perimental study was also conducted after the compressed scheme realized on the frame of JPEG2000. It is proved that the method of image compression based on region of interest automatic detected was reliable effective, and significant in applications.

关 键 词：红外图像 感兴趣区 图像压缩 JPEG2000 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]