哈尔滨工程大学学报2015，Vol.36Issue(7)：992-996,5.DOI:10.3969/j.issn.1006-7043.201404087
基于StOMP稀疏方法的高光谱图像目标检测
Hyperspectral image target detection algorithm based on StOMP sparse representation
摘要
Abstract
This paper proposes a new hyperspectral image ( HSI) target detection method using StOMP reconstruc-tion algorithm. StOMP could be used for the case when the computing cost of traditional sparse detection algorithms is very high because the HSI often has a large amount of data. The sparse representation algorithm has been success-fully applied to the field of HSI target detection and it has achieved good results. The new method improves the step of solving sparse coefficients, reducing the number of iterations of this process, which significantly improves the de-tection efficiency and reduces the computing cost. There are two sets of data in the simulation experiment and the results showed that the StOMP algorithm improves the detection speed and precision effectively.关键词
高光谱图像/目标检测/稀疏表示/StOMP算法/快速运算Key words
hyperspectral image/target detection/sparse representation/StOMP algorithm/fast computing分类
信息技术与安全科学引用本文复制引用
赵春晖,靖晓昊,李威..基于StOMP稀疏方法的高光谱图像目标检测[J].哈尔滨工程大学学报,2015,36(7):992-996,5.基金项目
国家自然科学基金资助项目(61405041) （61405041）
黑龙江省自然科学基金重点资助项目( ZD201216 ) （ ZD201216 ）
哈尔滨市优秀学科带头人基金资助项目( RC2013XK009003) . （ RC2013XK009003）
