论文部分内容阅读
为了及时监测溢油事故相关信息,防止污染扩张,该文利用2010年4月29日、5月17日及5月24日3个时相的中分辨率MODIS数据,以2010年发生的墨西哥湾溢油事故为例进行目标提取与识别。提取算法考虑到油膜和背景海水易混分的现象,采用了辅加纹理特征量的光谱角匹配算法进行油膜提取;同时,对比了常用的最小距离和支持向量机分类方法;并基于混淆矩阵对结果进行精度验证。结果显示,改进的光谱角匹配算法可以更准确地提取海面目标,精度高达90%以上,能够较好地监测出大面积溢油,可以用于海面溢油灾害的动态监测。
In order to timely monitor the information related to oil spill accidents and prevent the pollution from expanding, this paper uses MODIS data from three phases of April 29, May 17 and May 24, 2010 to analyze the Gulf of Mexico An example of oil spill extraction and identification of the target. The extraction algorithm takes into account the phenomenon that the oil film and the background sea are easy to be mixed and segmented. The spectral angle matching algorithm with additional texture features is used to extract the oil film. At the same time, the commonly used minimum distance and support vector machine classification methods are compared. Based on the confusion matrix The results of precision verification. The results show that the improved spectral angle matching algorithm can extract the sea targets more accurately, with the precision as high as 90%. It can monitor large area oil spill and can be used for the dynamic monitoring of sea surface oil spill disaster.