Abstract

The video tracking method based on single template matching is affected by the shape and transformation of single template, resulting in large tracking error and poor tracking effect. In this paper, a method of object tracking in sports video based on multi template matching is proposed. After preprocessing the video image, the background difference is used to detect the target in the video. Based on the selected benchmark template, the matching template is generated by multi-level geometric transformation and the template updating strategy is formulated. Extract the features of the template and image, and complete the tracking of sports video target by calculating the matching degree between multiple templates and image features. Through the contrast experiment, it is proved that the multi template matching tracking method has the advantages of short tracking time, high precision and better tracking effect.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call