Experimental Study of Intelligent Lighting Control Method for Dark Field Surveillance

A kind of intelligent lighting control method for the dark field surveillance is proposed. The design of it includes an offline stage and an online stage. Regarding the offline stage, first a visual ergonomic experiment is used to accumulate image datasets which have different subjective image quality evaluation degrees (IQEDs) for the typical surveillance application. Second, the objective IQED metrics are computed for these datasets above: the image region contrast, the image edge blur, the image gray deviation, and the image noise. Third, the k-means cluster method is employed to analyze the distribution thresholds of the objective metrics. Regarding the online stage, first the objective IQED metrics are computed for the input image. Then, the computed objective results will be compared with the distribution thresholds gotten in the offline stage. Finally, a kind of optimal lighting control will be performed. A lighting experimental system is built, and many experimental results have verified the correctness of the proposed method.