Improvement of top-hat based on soft morphology
基于软形态学的top-hat算子改进算法

To detect infrared small target under complex background, this paper presented a new class of top-hat operator based on soft morphology, and proposed a background suppression method based on soft morphology. In the proposed framework, it used mutual information to select the structure element scale adaptively, divided the structuring element into two parts, i. e. , the core and the soft boundary in order to meet the actual situation of the real target. Finally, designed a novel filtering framework, called balance structure to suppress filtering bias. Experiment results show that the output SNR of complex sea clutter background with 2. 36 dB can achieve 22. 51 dB. In the sense of visual effect and quantitative index, the proposed method performs better than several mathematics morphology based methods such as the classical top-hat, modified top-hat, soft top-hat and suppresses the infrared images of the complex background effectively.