Infrared Ship Detection Technology Based on Improved Residual Spectral Significance Detection
In order to solve the problem that the traditional significance detection method is not accurate in infrared ship target detection,a significance detection method based on improved residual spectrum is proposed.Firstly,a-daptive hexagon median filtering and wavelet image fusion are used to remove noise from infrared images.And then,based on the residual spectral significance detection algorithm,superpixel segmentation is introduced to process the target area of the ship to reduce the amount of calculation,while retaining the edge information of the ship target.In addition,the target mask is introduced to eliminate background interference to improve the accuracy of the algorithm.Finally,the detection of ship targets is realized by extracting a salient map of the image.Experi-ments show that this approach can achieve an accuracy about 3％higher than that of the previous,and therefore can help to accurately detect ship targets in complex backgrounds such as islands.
infrared ship targetadaptive hexagon median filteringsignificance detectionsuperpixel segmentation
国家科技期刊平台
NSTL
万方数据
