PSF估计与随机森林特征增强的超分辨率重建  

PSF estimation and super-resolution reconstruction of random forest feature enhancement

作　　者：郑艳 白旭 金鑫[3] 陈克海 ZHENG Yan;BAI Xu;JIN Xin;CHEN Kehai(Department of Information Engineering in Surveying Mapping Science and Remote Sensing,Guangdong Polytechinc of Industry and Commerce,Guangzhou 510510,China;Sinopec Petroleum Engineering Corporation,Dongying 257000,China;Beijing Urban Construction Exploration&Surveying Design Research Institute Co.,Ltd.,Beijing 100101,China)

机构地区：[1]广东工贸职业技术学院测绘遥感信息学院,广州510510 [2]中石化石油工程设计有限公司,山东东营257000 [3]北京城建勘测设计研究院有限责任公司,北京100101

出　　处：《测绘工程》2023年第6期60-67,共8页Engineering of Surveying and Mapping

基　　金：广东省教育厅2021年度广东省普通高校特色创新类项目(2021KTSCX207)。

摘　　要：光学遥感影像在形成、传输和记录过程中存在模糊、失真、噪声等降质现象,且图像缺少细节信息,提出一种基于PSF估计和细节特征增强的多帧超分辨率重建方法。针对POCS方法重建的影像纹理细节信息、边缘保持不足的问题,采用比值稀疏约束模型估计影像的点扩散函数,采用Lucas-Kanade(LK)光流法配准,利用随机森林方法进行边缘特征、细节特征的增强。以序列“珠海一号”影像为数据源,进行实验分析。结果表明,文中方法的信息熵可提升0.06~0.5 dB,平均梯度提升0.03~2.0 dB,提高重建影像的纹理细节信息和边缘清晰度,可获得更高质量的数据源。There are degradation phenomena such as blurring,distortion,and noise in the formation,transmission,and recording of optical remote sensing images.The lack of detail information in the image affects the later application of the image.This paper proposes a multi frame super resolution reconstruction method based on PSF estimation and detail feature enhancement.To solve the problem of insufficient texture detail information and edge preservation in images reconstructed using POCS methods,firstly,a ratio sparse constraint model is used to estimate the point spread function(PSF)of the image,and Lucas Kanade(LK)optical flow method is used for registration.Random forest method is used to enhance edge features and detail features.Experiments were conducted using the serial“Zhuhai-1”image as a data source.The results show that the information entropy of this method has increased by about 0.06 dB to 0.03 dB,and the average gradient has increased by about 0.5 dB to 2.0 dB.This method has improved the texture detail information and edge retention ability of the reconstructed image,and got can provide higher quality data sources for applications.

关 键 词：超分辨率重建 凸集投影 随机森林 图像边缘 

分 类 号：P237[天文地球—摄影测量与遥感]