Joint sparse representation of complementary components in SAR images for robust target recognition

Abstract

An automatic target recognition (ATR) method of synthetic aperture radar (SAR) images is proposed by joint sparse representation (JSR) of the complementary components from the original SAR image. The shadow and target image are generated from the original image. A simple but effective segmentation algorithm is designed to separate out the shadow region. By replacing the shadow region with randomly selected background pixels in the original image, the target image is generated. Afterwards, the two components together with the original image are jointly classified based on JSR. Due to the extended operating conditions (EOCs) in SAR ATR, the shadow or target region may be corrupted. In this case, the sole use of the original image may bring some interference caused by the corruption. As a remedy, the joint use of the three components can effectively improve the robustness of the ATR method to various EOCs by complementing each other. To quantitatively evaluate the proposed method, experiments are conducted on the moving and stationary target acquisition and recognition dataset under various conditions. The results demonstrate the effectiveness and robustness of the proposed method.

Keywords:

• Synthetic aperture radar (SAR)
• automatic target recognition (ATR)
• complementary components
• joint sparse representation (JSR)

Additional information

Funding

This work was supported by National Key Basic Research Program of China [grant number 2014CB046300].

Notes on contributors

Shuguang Miao

Shuguang Miao received his MSc degree from China University of Mining and Technology in 2010. Currently, he is studying for his doctor's degree in China University of Mining and Technology and a lecturer at Huaibei Normal University. His present research interests include compressed sensing, Coal-Rock interface recognition and IOT of mine.

Xiaowen Liu

Xiaowen Liu received her PhD from China University of Mining and Technology in 2009. Currently, she is a professor at China University of Mining and Technology. Her present research interests include wireless sensor network, IOT of mine and Coal-Rock interface recognition.