Research On Cartoon Method Of Video Face

Video face cartoon is the research hotspot in the field of computer graphics in recent years, its related products not only brings a lot of fun for people’s entertainment life but also has a broad market prospect. At present there are a lot of kinds methods of generating video cartoon by computer, in this paper, we propose a method for generating video cartoon with face expression by combining the advantages of the mainstream methods. The method can design personalized cartoon image and correct the defects through human interaction, facial expression information getting by tracking real face in the video is used for driving the cartoon image to get the video cartoon, which has the same expression as the real face in video. On this basis, a method was proposed to generate a multi-style video cartoon through human interaction. This paper have further studied the method for generating multi-style video cartoon, i.e., how to generate the corresponding video cartoon against a multi-style video.We propose solutions for slow processing speed and low similarity of cartoon and face. The detailed research content is as follows:1. We proposed two video cartoon synthetic such as the optical flow and the feature point method. Based on the optical flow, the first algorithm get facial motion data by estimating optical flow of video frames, which drive the face model to generate the video cartoon. The second algorithm, which is based on the feature points, analyze the movement of facial organs according to the location of the feature points, and generate the video cartoon by driving each facial organs.2. In order to generate more personalized video cartoon, a method was proposed on the basis of video cartoon synthetic to generate a multi-style video cartoon. We launch the research from two aspects, hair replacement and glasses addition, and put forward the corresponding algorithm for each rendering. It has shown that the proposed algorithm can generate multi-style video cartoon for each operation.3. A method for generating video cartoon of multiple angle was proposed. In this paper, algorithm research was conducted on the estimate of the swing angle, and proposed the corresponding algorithms. Experimental results show that this method can generate video cartoon of multiple angle which is consistent with real swing angle.4. The software implementation of the video cartoon generateing system was also given to complete all the content of the above algorithms. Through the system proposed in this paper, the users can get their own video cartoon eseasily.