Background:
The existing methods based on infrared patch image (IPI) model for small target detection does suffer from l1 norm sparsity problem where the non-target elements in the background image may sometimes be considered as the target element. Hence using l1 norm may lead to degrade the detection ability of small and dim target in a noisy environment. So a robust method needs to be developed to tackle the above-said problem.
Method:
In this paper, the Infrared patch model based on non-convex weighted nuclear norm minimization and Robust principal component analysis (IPNCWNNM-RPCA) is presented. Here we improve the existing IPI model by replacing the nuclear norm by weighted nuclear norm, where unlike in nuclear norm minimization we assign different weights to singular values. The proposed method is further solved by the alternating direction method of the multiplier (ADMM).
Results:
To validate the robustness of the proposed method extensive experiments on the large dataset is performed and the results indicate that the proposed method has responded shown good result against the other state-of-the-art methods.
Conclusion:
This paper presents a robust method in the name of infrared patch image model based on non-convex weighted nuclear norm minimization which improves the existing IPI based method for infrared dim and small target. The proposed method not only suppress the noise background nicely but also detect the target correctly.
SAR IMAGE DENOISING BASED ON MULTI-SCALE WEIGHTED NUCLEAR NORM MINIMIZATION
