一种基于SAR影像的绿潮分布信息提取方法  

A Green Tide Extraction Method Based on Markov Random Field Improved Algorithm of SAR Images

作　　者：宋清泉 王春晓[2] SONG Qingquan;WANG Chunxiao(Land Space Data and Remote Sensing Technology Center of Shandong Province,Jinan 250002,China;College of Oceanography and Space Informatics,China University of Petroleum,Qingdao 266580,China)

机构地区：[1]山东省国土空间数据和遥感技术中心,山东济南250002 [2]中国石油大学(华东)海洋与空间信息学院,山东青岛266580

出　　处：《海洋测绘》2020年第5期49-52,共4页Hydrographic Surveying and Charting

摘　　要：为基于SAR影像准确的提取海域中的绿潮范围,利用黄海地区2016年7月4日的RADARSAT-2卫星影像,分别运用马尔科夫随机场、最小核值相似区和局部二值模式3种算法提取绿潮并进行对比,提出了综合马尔科夫随机场和最小核值相似区算法。结果表明,3种算法中马尔科夫随机场算法提取效果最好,与实际情况最为接近,但改进的绿潮提取算法,提取效果更好,精度比单独使用马尔科夫随机场算法提高了18%。In order to extract the green tide range in the sea area accurately based on SAR image,a RADARSAT-2 satellite image of the yellow sea on July 4,2016 is used in this paper.It uses markov random field,smallest univalue segment assimilating nucleus and local binary patterns algorithm respectively,and then proposes a comprehensive algorithm of markov random field and smallest univalue segment assimilating nucleus algorithm after comparison.The results indicate that markov random field algorithm has the best extraction effect among the three algorithms,which is closest to the actual situation,but the improved green tide extraction algorithm has a better extraction effect and the accuracy is 18%higher than the markov random field algorithm.

关 键 词：合成孔径雷达 MRF算法 SUSAN算法 LBP算法 改进算法 绿潮提取 

分 类 号：P237[天文地球—摄影测量与遥感]