Lightness enhancement for video images based on nonlinear filter

It is difficult to enhance low-light video images, because of the irregularity of lightness for each frame and due to the length of execution time required for enhancement. In this research, the videos were optimised using a method based on the development of the technique of the histogram equalisation by using the sigmoid function as a non-linear conversion of the light function in the HSV colour space, this algorithm called Adapted Histogram-Equalization based on Sigmoid Function (AHSF). In order to know the efficiency of the enhancement of the videos, the proposed method was compared with several other methods including traditional as Histogram Equalization (HE) and Multi-Scale Retenix with Color Restoration (MSRCR) non-traditional as Fusion at Weakly Illuminated Images (FWII), This comparison was made by calculating the entropy rate for each video as well as calculating the execution time. By analysing the results, it was found that the proposed algorithm AHSF has good results in improvement and less implementation time.