An improved method for dim infrared target detection with nonsubsampled Contourlet transform

For the problem of small infrared target detection with complex background interference and noises, an improved dim small infrared target detection algorithm based on nonsubsampled Contourlet transform (NSCT) is put forward. Firstly, the infrared image with dim small target is preprocessed. Then, the NSCT is implemented, by combining the improved nonlinear map function and energy cross fusion to suppress the background clutters. Finally, threshold segmentation using the Otsu algorithm is used to isolate the dim small infrared target.compared with the similar small target detection algorithm, the method can effectively detect the small infrared target.