Abstract
The color scheme of interactive graphic design often needs to draw inspiration from the color styles of some images during the design process. Flat images are an important conceptual source for product color schemes. Image matching has become an indispensable technology in modern information processing and is also widely used in other fields. Therefore, conducting research on relevant algorithms in image matching technology has profound significance for the development of computer vision. This article optimizes the color matching algorithm for interactive graphic design based on RL (Reinforcement Learning). The research results show that the average matching accuracy of GA (Genetic Algorithm), CF (Collaborative Filtering), and our algorithm for each image is 86.24%, 90.47%, and 95.02%, respectively. From the test results, it can be found that the accuracy of the algorithm in this paper is the highest among them. From Figure 5, it can be seen that the average matching time of the three algorithms: GA, CF, and our algorithm for each image is 27.51, 24.37, and 6.54, respectively. From the test results, it can be found that the average matching time of the algorithm in this paper is the shortest. RL can effectively express the global structural information between image pixels.