Light Stretch Algorithm for Image Quality Enhancement

Images taken from outdoor scenes are usually degraded due to weather condition, resulting in reduction in color fidelity as well as contrast of captured images. This gives rise to difficulties when processing these images by an algorithm designed for clear weather condition. In lieu of refining individual algorithms to be bad weather-appropriate, it sounds more reasonable to perform on each input image an enhancement step. Thus, image quality enhancement is currently in great demand. In this paper, we hereby propose the light stretch algorithm, which simultaneously performs intensity expansion and reduction by solving the saturation phenomenon of bright color region and the discoloration of dark color region in an image. The simulation result strengthens its superiority over the other two methods based on unidirectional image enhancement and Photoshop curve adjustment method.