Awesome

Code-SRWS

This repository is for Self-Regularized Weighted Sparse (SRWS) model introduced in the following paper and is built in Matlab R2018a.

Zhang T, Peng Z, Wu H, et al. Infrared small target detection via self-regularized weighted sparse model[J]. Neurocomputing, 420: 124-148.

For more information about me, you can visit my persional website.

Contents

1. Introduction
2. Method
3. Test
4. Results
5. Citation

Introduction

Infrared search and track (IRST) system is widely used in many fields, however, it’s still a challenging task to detect infrared small targets in complex background. This paper proposed a novel detection method called self-regularized weighted sparse (SRWS) model. The algorithm is designed for the hypothesis that data may come from multi-subspaces. And the overlapping edge information (OEI), which can detect the background structure information, is applied to constrain the sparse item and enhance the accuracy. Furthermore, the self-regularization item is applied to mine the potential information in background, and extract clutter from multi-subspaces. Therefore, the infrared small target detection problem is transformed into an optimization problem. By combining the optimization function with alternating direction method of multipliers (ADMM), we explained the solution method of SRWS and optimized its iterative convergence condition. A series of experimental results show that the proposed method outperforms state-of-the-art baselines.

Method

The iterative process of the SRWS model is given in the following tables.

Test

Quick start

1. Download the code and test images in ./TestCode/.

2. Modify the image path in the demo.m, and run the file.

Results

Visual Evaluation

Multi-Scene Adaptability

Visual Comparison with Baselines

Quantitative Evaluation

ROC

To better compare the AUC of each of the curves in Figure 10, their specific values are listed in the following table, where the maximum value of each sequence AUC is indicated in red and the second largest value is indicated in purple.

SCRG and BSF

Iteration Number and Running Time

For more information, please refer to our paper

Citation

If you find the code helpful in your resarch or work, please cite the following papers.

Zhang T, Peng Z, Wu H, et al. Infrared small target detection via self-regularized weighted sparse model[J]. Neurocomputing, 420: 124-148.

@article{zhang420infrared,
  title={Infrared small target detection via self-regularized weighted sparse model},
  author={Zhang, Tianfang and Peng, Zhenming and Wu, Hao and He, Yanmin and Li, Chaohai and Yang, Chunping},
  journal={Neurocomputing},
  volume={420},
  pages={124--148},
  publisher={Elsevier}
}