Goaling recognition based on intelligent analysis of real-time basketball image of Internet of Things

This research aims to study the goaling recognition methods based on the intelligent analysis of real-time basketball images of the Internet of Things, so as to accelerate the development of basketball and other sports fields. First, the status of basketball-related image recognition and tracking in the field of sports is analysed, which are combined based on the background differencing and inter-frame differencing. Then, an improved image intelligent detection is proposed to identify the basketball goaling, and further analysis of the tracking algorithm related to the moving target is conducted. Finally, an improved Camshift target tracking algorithm is proposed to detect and recognize the basketball goaling. The results show that both background differencing and inter-frame differencing have large errors in basketball goaling recognition. The improved image intelligent detection algorithm in this research can fit the contour of the basketball very well, and the fitting effect of the smallest circle is ideal. From the perspective of basketball goaling trajectory tracking, it is found that the improved Camshift goal tracking algorithm has the best recognition effect on basketball goaling. In addition, the average running time of the algorithm is 209 ms, with favourable accuracy. Therefore, verified by the tests, the intelligent detection algorithm and trajectory tracking algorithm proposed in this research have high recognition accuracy for basketball goaling, and the operation efficiency is ideal, which provides experimental reference for the later intelligent development of sports fields such as basketball.