基于图像流的近实时遥感图像在线目标检测方法  

Near-real-time remote sensing image online target detection algorithm based on image flow

作　　者：陈金勇[1] 王敏 高峰[1] 孙康[1] 许妙忠[2] Chen Jinyong;Wang Min;Gao Feng;Sun Kang;Xu Miaozhong(The 54th Research Institute of China Electronics Technology Group Corporation, Shijiazhuang 050081;State Key Laboratory of Information Engineering in Surveying, Mapping, and Remote Sensing, Wuhan University, Wuhan 430079, China)

机构地区：[1]中国电子科技集团公司第五十四研究所,石家庄050081 [2]武汉大学测绘遥感信息工程国家重点实验室,武汉430079

出　　处：《国外电子测量技术》2018年第8期72-76,共5页Foreign Electronic Measurement Technology

基　　金：2017年河北省“三三三人才”工程项目(A2017010001)资助

摘　　要：随着卫星遥感技术的快速发展,多源、高分辨率遥感数据呈爆炸式快速增长,遥感数据的存储和处理方面遇到了前所未有的挑战。为了能够从海量的遥感信息中快速准确的提取出有效的目标信息,提出了一种基于图像流的近实时遥感图像在线目标检测方法,该方法能够在进行图像生产的同时进行图像目标检测。同时,该方法也适用于生产后图像的分块处理。该方法给出了一种自适应的图像分块方法,根据图像尺寸、分辨率、目标大小及目标检测算法自动确定图像分块处理的大小,从而使得总体计算复杂度最低。使用CFAR目标检测算法对该方法进行了仿真验证,实验结果表明,单线程下,对大小为12 000×18 405,分辨率0.6m的图像,本方法处理时间为35.81s,大大降低了计算复杂度,有效提高了计算效率。With the rapid development of satellite remote sensing technology,multi-source and high-resolution remote sensing data is exploding,so the storage and processing of sensing data encounter unprecedented challenges.In order to extract effective target information quickly and accurately from massive remote sensing information,it proposes a near real-time online target detection method based on image flow for remote sensing images,which can produce image and detect targets simultaneously.This method can be also used for accelerating process progress of high-level remote sensing imagery products by image blocking.This method gives an adaptive image segmentation method,and automatically determines the size of image segmentation based on image size,resolution,target size and target detection algorithm,so that the overall computational complexity is the lowest.It uses CFAR target detection algorithm to verified by simulation with this method and experimental results show that under the single thread,the processing time is 35.81 swith the size12 000×18 405,resolution of 0.6 meters,greatly reduces the computation complexity and improve the calculation efficiency.

关 键 词：遥感图像 目标检测 图像分块 在线检测 

分 类 号：P237[天文地球—摄影测量与遥感]