Golf Swing Detection for Video Indexing in Golf Practice Video

• Main Authors: HUANG, SHANG-MING, 黃上銘
• Other Authors: Lien, Brian Kuo-Jen
• Format: Others
• Language: zh-TW
• Published: 2015
• Online Access: http://ndltd.ncl.edu.tw/handle/06254168485494279091

碩士 === 輔仁大學 === 資訊工程學系碩士班 === 103 === This thesis presents a new algorithm for golf swing video indexing that analyzes and track the golf club head from the golf practice video. The Golf swing video is the most important material for golfer because it shows golfer's swing problems. Golf swing is distributed in the following four stages: address, takeaway, backswing and impact. When the golfer swings the club in the stages of address, takeaway, backswing and impact in order we call this swing a complete swing. To detect a complete swing, first the golf club head is detected using temporal difference and Morphological processing, then the four stages of swing are recognize by calculating the distance between the object center and frame center. The experimental results show that the algorithm can efficiently extract the golf swing from the practice video and has better performance then the original algorithm.