基于小波分析的海洋锋形态特征提取

遥感影像中海洋锋具有弱边缘(Weak Edge)信息特性,其空间尺度在不同的空间位置存在较大的差异。针对传统边缘算子只能提取某一尺度的海洋锋信息,而损失其他尺度信息不足,本文利用小波多尺度分析特性,探讨了海洋锋信息在各尺度间的分布的继承性,提出最佳尺度确定的算法框架,设计出最佳尺度确定算法的流程,进一步进行海洋锋各尺度信息的提取。最后以海面温度锋-墨西哥湾流为实例进行形态特征提取,并与典型的边缘提取算子效果在信息量、抗干扰性、定位精度、连续性和对比度上进行对比分析。试验结果表明了该算法在提取多尺度的海洋锋信息上的可行性和实用性。

Extraction of Ocean Fronts Based on Wavelet Analysis

Ocean fronts are characterized by weak edges in remote sensing images, with the space scales having great difference in different space sites. Since the traditional edge operators can only extract some scale information, and the other information will be missed, the adaptive algorithm to define optimal algorithm is presented based on MRA. And the flow of the optimal algorithm is designed based on the framework. In the end, taking the Mexico Gulf for example, the temperature fronts are extracted, and their effects are compared with the traditional algorithm on information amounts, interference, positioning accuracy, continuity and contrast. The results show that it is practicable and feasible to extract the ocean fronts through the optimal scale adaptive algorithm based on Mallet MRA.