The tracking study of 3D image feature points of aerobic movements of high difficulty

摘要

Tracking the 3D image feature points of aerobic movements of high difficulty can improve the quality of aerobic movements. The current method takes the windowed area as the tracking template to complete the track of feature points with low efficiency. For that reason, a method based on Gabor is presented with which to track the 3D image feature points of aerobic movements of high difficulty. The method extracts feature points with SIFT algorithms, classifies feature points according to AdaBoost algorithms, and tracks feature points through the image pyramid. The experiment shows that the mentioned method enhances the accuracy of tracking the 3D image feature points of aerobic movements.