Abstract
Abstract At present, the method based on constant false alarm rate (CFAR) technology is one of the important means of marine target detection. The performance of traditional sliding window CFAR detectors is easy to be affected by speckle noise and multi-target crosstalk, resulting in a low success rate of visual inspection. In this paper, a CFAR detection method based on superpixel difference degree is proposed. This method can effectively distinguish potential target superpixels and background superpixels, and then through the selection of background superpixels, the clutter probability density function model is fitted with the unbounded Johnson distribution model, to carry out CFAR detection. Based on the SAR images of the Gaofen-3 satellite and Sentinel-1 satellite, the experiment verifies that the method can improve the detection performance in multiple scenarios.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call