Image Processing Algorithm In Image Capture System

With the increasing incidence of the night, the requirement is higher than before for the night vision surveillance systems in image capturing. It is needed to improve night monitoring results. A color transfer technology is made in this paper. An algorithm is proposed for color transfer based on extraction of color feature from the block thinking, the second method is based on the Gray Level Co-occurrence Matrix in the extraction of texture feature, and the last one employs technology of color image fusion and restoration. The gray image is colorized.Colorization of gray image is to pass the color characteristics of a color image to another piece of gray-scale image, so that the source grayscale target image is colorized with similar color as the color one. This algorithm is proposed to resolve the problem of the traditional color transfer method, which has the poor adaptive ability. Automatic matching is completed in the source image and destination one. Image color and texture features are calculated in order to find images that match the target image as a source of color transfer images in the image library. The adaptability of the algorithm is improved. Artificial participation is reduced in the process of color transfer. The efficiency and effect of transmission of color are all improved. Two algorithms are proposed, one is extraction of color feature based on block thinking, and the other is extraction of texture feature based on GLCM. According to the characteristics of night vision images, the color transfer effect is not very satisfactory if color images are passed directly from the night-vision color of gray image. In this case, the dim light and infrared images are fused from the collected images, so the clarity of the target image is improved. The target image is selected from the fused images, and process of colorized is made for the target one. The matching source image is found for the target one by calculating. The color transfer is finally completed.The colors transfer software is designed in the platform of VS2008. Practicability is verified through simulation of many images for the proposed algorithm of color transfer.