Automatic Detection Algorithm of Ocean Mesoscale Eddies Based on SAR Images
Mesoscale eddies in the ocean are a common and important marine phenomenon,and currently synthetic aperture radar technology(high-resolution imaging of the sea surface)is used to provide conditions for observing and studying scale eddies in the ocean.Due to the complexity and diversity of mesoscale vortex morphology,it is very difficult to detect mesoscale eddies.To solve the problem of efficient detection of mesoscale eddies,an automatic detection algorithm for mesoscale eddies in the ocean is proposed.The algorithm uses"patrol line extraction"to extract and cluster the texture of high-resolution sea surface images,transforming the texture features in the images into a set of line segments that describe the direction of the ocean flow field.Then,the geometric features of the line segments are combined with the laws of eddies to calculate the mesoscale vortex features,and the detection of mesoscale vortex centers is achieved through threshold screening.This algorithm has strong flexibility and rich functionality,which can help researchers obtain sufficient samples for studying mesoscale eddies and has high application value.GF-3 and Sentinel-1 satellite images were selected for the experiment.The results indicate that the proposed detection algorithm can accurately detect mesoscale eddies and visualize them.
SAR ocean remote sensingmesoscale eddiesautomatic detection
万方数据
维普
