Fig. 10From: Machine learning hyperparameter selection for Contrast Limited Adaptive Histogram EqualizationLB-CLAHE’s low performance according to the IQA techniques. Best results achieved by global HE (A) and (B). Best results achieved by fixed CLAHE (C) and (D). Ideal image (1), distorted image (2), global HE (3), fixed CLAHE with CL=0.01 and NT=[8,8] (4), proposed LB-CLAHE (5)Back to article page