An Adaptive Infrared Image Preprocessing Method Based on Background Complexity Descriptors

Infrared image complexity is applied widely in target recognition of complex scenario, infrared imaging system performance prediction and evaluation, performance comparison of target recognition algorithms, and other fields. This paper presents an adaptive infrared image preprocessing method. First, the variance weighted information entropy for quantitative description of Infrared background complexity is proposed. Second, the optimal cut-off frequency of the filter is designed by the background complexity of the real-time image. Finally, an improved high-pass Butterworth filter is used for adaptive preprocessing for complex infrared images