Image processing analysis and research based on game animation design

With the development of the game industry, and the new electronic consumption chain to promote the development of the game industry model continues to strengthen. Therefore, it is more necessary to make good use of good technology to do a good job of games. From the momentum of the development of mobile terminal games in recent years, the game industry has become a new economic growth point in China's economic transformation. With the development of computer and Internet, graphics and image processing technology has entered another unprecedented stage. In recent years, different image processing techniques have been used to process and analyze the game interface. Appropriate image processing methods include image enhancement, image binarization, image edge detection and image feature extraction. In this paper, by changing the mapping function from priority to probability and comparing the single mapping function of the previous algorithm, the mapping function of playback learning with higher probability of the important priority playback unit is found. In the experiment, the intuitive model strategy analysis of the improved algorithm is carried out firstly. Then the choice of CNN network layer structure, cost function analysis, efficiency analysis and game score comparison of each algorithm are carried out. Finally, the test results show that the new algorithm in this paper can make more effective decisions in video games and achieve the goal of winning higher scores and spending less time.