Night image enhancement technique using hybrid of good (high intensity) & poor (low intensity) images

Night image enhancement has been a subject of interest of various sectors. Border patrol, automotive, industry, defense are just a few examples. Accidents at night could potentially prove fatal. Thus, lately there is a significant interest from automotive industry in conducting research in night image enhancement. This paper presents an efficient night image enhancement technique using the hybrid image form bright images and poor light intensity images. We have several night image enhancement techniques to get better results but all of that are based upon filters and DSP based algorithms so will not get improved results. We implement a new and better way to enhance the quality of fixed camera on highways to indentify vehicle easily in very low intensity. Hybrid image can be obtained with the combination of fixed image (mean of several images in day) with the moving object (object in motion in night) with a simple algorithm. Need to obtain mean image for day is to get the final image in which there are only still objects in that area capturing by highway camera. Proposed algorithm enhances the night image visual quality by 80 percent with this hybrid model.