Adaptive CLAHE image enhancement using imaging environment self-perception

Abstract

A novel Contrast Limited Adaptive Histogram Equalization (CLAHE) image enhancement method which uses the environment self-perception mechanism is proposed. First, the typical degraded image datasets are collected. Second, several Image Quality (IQ) evaluation metrics are used to assess the imaging effect of these datasets above. Third, a BP network is employed to build the connection between the IQ evaluation results above and the optimal control parameters tuning results of the classic CLAHE. The optimal control parameters tuning results are gotten by the subjective evaluation and the setting of the control parameters of the classic CLAHE. The expert experiences of the optimal enhancement are used as the evaluation benchmark. Finally, when a new degraded image is captured, its IQ evaluation metrics will be computed and its optimal control parameters will be forecasted by the BP network and the computed IQ evaluation metrics. Many experiment results have shown the effectiveness of proposed method.