The Research Of Video Enhancement For Low Illumination

The image is low-light and has more noise captured by electronic equipment inlow illumination environment, so the visual effect is poor.The low-light videoenhancement technology is divided into two main parts:1.The skill of brightnessenhancement in low illumination environment;2.The denoising technology in lowillumination environment.The difficult of brightness enhancement is not losing theoriginal image’s details while enhancing the brightness. The difficult of denosing inlow illumination environment is that maintaining the image details and edge whiledenoising. To solve the above problem, this paper carries out the following work:1):Putting forward the filter based on the principles of spatial-continuity.The filtercontains two important criteria:the depth consistency criteria and spatial continuitycriteria which can solve the problem of estimating the same or different depth of aircurtain. Then we successfully applied the filter to image defogging while theprccessing speed is faster than the other method of image defogging at the same timeensuring the defogging effect.2):Putting forward the low illumination brightness enhancement technique basedon the principle of negative conversion. The algorithm use the last1) filtering toestimating the air curtain and maintain the image’s details and color when increasingthe brightness of the image rapidly. Finally the algorithm can be success-fully used inreal-time video surveillance system after optimization.3):Realizing the adaptive manifold filter which can remove the noise in lowillumination environment effectively at the time maintain the edge of the image. Thispaper makes the optimization and improvement of PCA and the Gauss function whichcalculating very complex.Finally the paper transplants the the enhancement algorithm of low lluminationinto the commercial streaming platform VLC，it can process720P video in reai-timeby making optimizing.