地理空间信息2024，Vol.22Issue(8)：81-84,4.DOI:10.3969/j.issn.1672-4623.2024.08.018
结合先验概率估计的SAR影像水体提取
Water Extraction from SAR Images Combined with Prior Probability Estimation
摘要
Abstract
Based on the SAR image and DSWE algorithm,we improved the existing threshold method,and proposed a method combining proba-bilistic distribution map with threshold value.We took DSWE water probability data set as the prior data set at first.And then,we divided the da-ta set into three categories by water body ratio,such as high,medium and low,and took it as the guiding threshold.We preprocessed the Senti-nel-1 data by SNAP software,classified the obtained backscatter coefficients according to the proportion of water in the prior data set,and used the bimodal histogram method,object-oriented method and morphological expansion to process the image and extract water.Finally,we used the simultaneous high-resolution image as the base image for visual interpretation.The experimental result shows that this method can extract the flood inundated area of Poyang Lake with high accuracy,which further proves the effectiveness and reliability of this method in water extraction.关键词
洪涝灾害/Sentinel-1/DSWE/水体分布数据集Key words
flood disaster/Sentinel-1/DSWE/water distribution data set分类
天文与地球科学引用本文复制引用
王泽雨,李建章..结合先验概率估计的SAR影像水体提取[J].地理空间信息,2024,22(8):81-84,4.基金项目
国家自然科学基金资助项目(41661091) （41661091）
兰州交通大学优秀平台资助项目(201806). （201806）
