A ‘divide and conquer’regularization imaging method for forward-looking scanning radar azimuth super-resolution

作者：Tan Ke; Li Wenchao; Huang Yulin; Zhang Qian; Yang Jianyu

来源：Progress in Electromagnetics Research M, 2018, 66: 151-161.

DOI：10.2528/PIERM18011005

• Divide and conquer
• Effective approaches
• Gradient informations
• Regularization methods
• Segmentation methods
• Sparse regularizations
• Super resolution
• Synthetic and real data

Sparse regularization imaging method (SRIM) is an effective approach to implement azimuth super-resolution for forward-looking scanning radar. However, for the scene that contains adjacent strong targets in the continuous weak background, SRIM may destroy the structure of the scene when trying to separate the closely located targets. In this paper, a divide and conquer regularization imaging method (DC-RIM) is proposed to solve this problem. Firstly, the data are divided into two channels by the mean-variance segmentation method. Normally, we consider that the data of channel I contain strong scatterers and that the data of channel II contain weak background. Afterwards, SRIM is conducted on channel I to distinguish the targets. For the data of channel II, a region enhancement regularization method is particularly proposed to acquire a good structure of the scene by making use of two-order gradient information of the data. Finally, a good imaging result can be obtained by combining the results of two channels. Experiments based on both synthetic and real data are given to verify the effectiveness of the method.

• 出版日期2018
• 单位电子科技大学