Background suppression for dim small target with Gabor kernel non-local means

Background suppression is one of the most important subjects in the field of low signal-to-noise ratio and low contrast dim and small targets detection from a sequence of infrared images.Due to the cluttered clouds and structured background,it is difficult to discern the dim and small targets in infrared dim and small targets detection.To improve the detection performance of dim and small targets from complex background,the background suppression method based on Gabor kernel non-local means was presented,and local structure tensor was introduced to suppress background details and enhance target information.The local structure tensor could reduce computational complexity and enhance real-time performance of the method.Compared with wavelet filtering and Gaussian-kernel non-local methods,the proposed method can suppress complex background and enhance the dim and small target effectively according to the visual effect and quantitative index obtained from several groups of experiments.