Research on target tracking technology based on machine learning

In recent years, a large number of scholars have been engaged in the research of target tracking algorithms, but target tracking is still a very challenging problem due to the variability of the observed target information in the tracking process, the mobility of the target and the complexity of the background. In this paper, relying on the theoretical basis of TLD tracking algorithm, implementation detection module, P-N learning module and synthesis module, the dynamic fusion features of the target in different states are used as target templates to take advantage of the different features of the target in different states and increase the tracking success rate of the algorithm. For the problem that the target motion background changes, when the target color is seriously affected by the background change or interfered by the similar target, Hog features are combined with color features to make the tracking algorithm track the target to the maximum extent. This study aims to set a new direction for research in this field, as a way to promote the update and iteration of the technology in this field.