Research Of Multiple Virtual Exposure Night Vision Image Contrast Enhancement Algorithm

At night, images have low contrast and information loss. It’s not convenient to capture detail information in many areas. In our daily life, digital cameras and smart phones have been widely applied in the photo shoot. But a digital camera and smart phone have a limitation in dynamic range. This limit is much lower than the scope that humans can sense. Therefore, in high dynamic range photos are often in shadows or areas of high brightness. These photos are underexposed or overexposed.In view of the present image recovery algorithms with good effects in the field of low illumination image enhancement, learn and simulate the dark channel prior theory and use this theory to enhance low illumination images. The shortcomings of dark channel prior are details missing and colors are slightly distorted. This paper proposes a improved dark channel prior night vision image enhancement algorithm. Transform the space of night vision images to RGB space and estimate that global atmospheric light and transmission coefficient. Optimize the transmission coefficient, and then use color keep to process the enhanced image ultimately capturing enhance images. The proposed new algorithm can maintain the edge and at the same time reduce the computational complexity. The specific details can be restored and the running time is also better.Then combine virtual exposure technology and image fusion technology. Propose a night vision image enhancement algorithm based on exposure fusion to enhance the images. First of all, input night vision images, using virtual exposure technology to produce mult i virtual exposure image sequences. Use image fusion technology to realize the image fusion of virtual exposure image sequences. Finally, process the fused image using color rendition to prevent relevant hue shift and color saturation decline, and consequently obtain the enhance images. The experimental results show that the algorithms proposed in this paper can effectively enhance the night vision images and improve image contrast as well as can highlight the details and reduce the noise.At the end of this paper, night vision image visibility evaluation indexes are put forward and an image sharpening system is realized. The system can deal with images under the environment of fog, low illumination, dust and underwater. The night vision image enhancement algorithms proposed in this paper as one of the modules can adjust enhancement strength and smoothness.