Research on Tennis Recognition Technology Based on Computer Vision

Visual attention mechanism is one of the important means for human beings to perceive the external world. Using mathematical models to introduce visual attention mechanism into computer vision to simulate human visual perception system is a hot research topic in the field of computer vision. The research of visual attention model is not only helpful for human beings to better explore the working mechanism of human visual attention, but also has very important significance for solving large-scale data screening and improving image processing efficiency, which has important application value in moving object detection, machine vision, image information matching, image compression and other fields. The visual attention model is used to preprocess the video sequence, and the region of interest is found as the candidate region for target detection. Then the color and shape matching algorithm is used to match the candidate regions. Experiments on tennis video show that the algorithm can recognize the target well when the color and shape of the target are relatively single and the significance in the background is high.