Adapted Single Scale Retinex Algorithm for Nighttime Image Enhancement

Images captured at night with low-light conditions frequently have a loss of visible details, inadequate contrast, low brightness, and noise. Therefore, it is difficult to perceive, extract, analyze important visual information from these images, unless they were properly processed. Different algorithms exist process nighttime yet most are highly complex, generate processing artifacts, over-smooth the or do not improve illumination adequately. Thus, single scale retinex (SSR) algorithm adopted in this study provide better for images. The proposed starts by converting color image RGB model HSV enhancing V channel only while preserving H S channels. Then, determined image’s illuminated version somewhat like SSR, computes logarithms original then subtracts two images utilizing an altered procedure. Next, modified gamma-adjusted Rayleigh distribution function applied, its outcome processed once more automatic linear contrast stretching approach produce that will be utilized preserved channels output image. developed assessed using real dataset evaluated three dedicated evaluation methods, compared ten dissimilar contemporary algorithms. obtained results demonstrated can significantly perceptual quality suppress artifact generation rapidly efficiently, addition showing ability surpass performance different existing subjectively objectively.