Detection and Tracking System for Long Jumping Skilled Videos

Abstract

Tracking the Long Jump Athletes in the video, trainers can more accurately analyze their action details during the entire jump, and the realization of identification and detecting of Long Jumpers in the video is essential for tracking athlete in technical video. Based on the Opencv vision library of the platform of Microsoft Visual Studio 2008, By calculating the probability distribution of the target pixel to positioning the athletes, the use of Kalman filter algorithm to predict its trajectory, to take a combination of both methods to achieve the recognition and detection of athletes.