An Algorithm of Film Image Coloring and Image Restoration Based on DL

In image processing, color is usually the most basic visual element, and it is the element that people can obtain the most effective information in cognition. It plays an important role in various fields. Black-and-white image coloring is an important part of image processing system. As a computer-aided technology, image coloring has a wide range of applications in film and television processing and old photo restoration. Because the algorithm does not have the ability to understand the movie content, it extracts the feature information of the image and maps it to the similar parts, thus completing the coloring work. According to the research, the human eye can only recognize dozens of gray levels of images, but compared with color images, the human eye can recognize thousands of colors. It can be seen that the influence of color images on human eyes is far greater than that of gray images. In addition, the gray-scale image is obviously not conducive to the feature extraction of the image information by the algorithm, so coloring and pseudo-colorizing the gray-scale image will be beneficial to the feature extraction and expression of the image. And the colored image can better express the meaning of the image, so that the human eye can capture the meaning and information of the image itself faster, better and more accurately, and then give full play to and express the value contained in the image.