Contrast enhancement algorithm for outdoor infrared images based on local gradient-grayscale statistical feature

Contrast enhancement for infrared images is important in various night vision applications. However, existing local contrast enhancement algorithms often over-enhance smooth regions in outdoor infrared images. To address this limitation, this paper presents a contrast enhancement algorithm based on local gradient-grayscale statistical feature. The proposed algorithm first extracts such features from image sub-blocks, then classifies the sub-blocks as either simple or non-simple based on textural complexity using a model trained by a support vector machine, and subsequently adopts different grayscale mapping strategies to process the two types separately. Experimental results show that the proposed algorithm avoids over-enhancing simple regions while effectively improving the contrast in regions with more details.