SAR image ship detection method based on abnormal detection and double-layer screening mechanism

Abstract

The SAR image ship detection method based on abnormal detection and double-layer screening mechanism comprises: a, carrying out a detection preprocessing stage, specifically, extracting a spectral vector, an SAR image is converted into a hyperspectral type image, then an interested target area is extracted through a hyperspectral anomaly detection algorithm, and the area comprises all areas whereships may exist; (1) performing image conversion: taking a sliding frame from an image to be processed according to pixels, and traversing to obtain converted image data consisting of spectral vectorsof each pixel of an original image, namely finishing image conversion. The method has the advantages that the SAR image is converted into the hyperspectral image, preprocessing of ship target detection is achieved through the hyperspectral image anomaly detection algorithm, the binary image of the interested area is obtained, and on the basis, accurate modeling of background clutters and rapid detection of ship targets are achieved through the double-layer screening mechanism.

Description

technical field [0001] The invention relates to a SAR image ship detection method based on anomaly detection and a double-layer screening mechanism, and belongs to the technical field of detection methods. Background technique [0002] Ship target detection is a key technology in the application of Synthetic Aperture Radar (SAR) images in sea area monitoring. At present, the constant false alarm rate (Constant False Alarm Rate, CFAR) method is still the most widely studied in this field, and it is also a relatively practical method. Among them, intelligent CFAR detection can be used in complex environments such as multi-target environments and non-uniform clutter areas. A good detection effect can be obtained. Most intelligent CFAR detection algorithms remove high-brightness target pixels in the background through a screening mechanism, and then perform background clutter parameter estimation, such as intelligent CFAR algorithms based on automatic screening, and target dete...

AI Technical Summary

Problems solved by technology

The former needs to determine the screening depth in advance based on prior knowledge, which is usually difficult to obtain the optimal value; the latter's ICS convergence speed is slow, and multiple iterations increase the amount of computation

These intelligent detection methods screen the pixels in the sliding frame to varying degrees in the clutter background modeling stage, but there are still some target pixels with low gray values ​​that are misjudged as the background and retained, which reduces the screening efficiency. Accuracy, leading to errors in parameter estimates

Images