高精度船舶高光谱图像的解混方法  

Unmixing method for high precision ship hyperspectral images

作　　者：黎瑾慧 LI Jin-hui(Henan Vocational College of Water Conservancy and Environment,Zhengzhou 450008,China)

机构地区：[1]河南水利与环境职业学院,河南郑州450008

出　　处：《舰船科学技术》2021年第20期202-204,共3页Ship Science and Technology

基　　金：河南省职业教育教学改革研究与实践项目(ZJB20260)

摘　　要：典型的高光谱图像解混利用端元选择技术,而高精度图像信息存在冗余,导致端元选择繁琐,影响了解混处理的效率和质量。因此,以提升光谱图像解混处理效率为研究目的,将研究高精度船舶高光谱图像的解混方法。利用改进连续投影算法对高光谱图像进行降维处理后,在生成对抗网络中引入Lipschitz约束和光谱规则,实现对高光谱图像的解混处理。仿真实验结果表明,在处理噪声为40 dB图像时,方法的处理时间仅为1.18 s,且相比其他方法,解混后图像的质量更佳。Typical hyperspectral image unmixing utilizes end element selection techniques,and the redundancy of high precision image information leads to tedious end element selection,which affects the efficiency and quality of unmixing processing.Therefore,with the research purpose of improving the efficiency of spectral image unmixing processing,the unmixing method of high-precision ship hyperspectral images will be investigated.After the hyperspectral image is downscaled by the improved continuous projection algorithm,the Lipschitz constraint and spectral rules are introduced into the generative adversarial network to achieve the unmixing process of hyperspectral images.The simulation results show that the processing time of the method is only 1.18 s when processing the image with 40 dB noise,and the quality of the unmixed image is better than other methods.

关 键 词：高精度图像 高光谱图像 光谱解混 图像降维 连续投影算法 生成对抗网络 

分 类 号：TP75[自动化与计算机技术—检测技术与自动化装置]