Novel Adaptive Thresholding Technique for Poor Lighting Document Image Based on Human Vision Characteristic
基于人眼视觉的对不良照明图像的二值化方法(英文)

An algorithm of adaptive threshold partition based on human visual characteristic was proposed. All the image blocks were divided into two types by classification algorithm based on human visual contrast sensitivity. Then blocks with different types were binarized by OTSU and Multi-scale fuzzy membership grade algorithm respectively. The algorithm was realized with VC++ language. Experimental results show that this proposed method is well suitable for poor quality document image binarization.