Leveraging Video Analytics to Revamp Golf Training and Performance Analysis

Sports data analysis equips one with the knowledge and abilities needed to incorporate a performance analysis method into coaching. Golf, a technique-based activity, is one of the most difficult sports to master. Finding the right swing, as well as the right clubs (irons and woods), is critical in obtaining top performance. Advanced movement accuracy is also required for putting, chipping, pitching, and bunker shots.

To deliver a modern video analysis system for analyzing golf swings for training the amateurs from professional golfers’ videos.

• Algoscale performed Video Analytics to create a model that could discern between different types of clubs used by a golfer in a video and then grade a player’s shot based on the trail.
• To get the trail of each shot, we calculated multiple angles of the club with the body in different frames. Then, to compare it to the shot of a novice player, a similar operation was performed on professional videos.
• Finally, the cosine similarity between the two trails was calculated to determine the final grade. Training data was extracted from progolfswingvideos.com and golfloopy.com. The entire video was broken down into distinct frames of imagery at various time intervals.
• Fast R-CNN algorithm was used to train the model, while computer vision techniques and complex mathematics were used to compare the two shots.

• Algoscale assisted in the development of a cutting-edge video analysis system for evaluating golf swings and instructing amateur golfers using professional golfers’ videos.
• Our solution would help the trainee in becoming a better player with time by allowing them to compare their shots to those of a professional golfer.
• Identification accuracy of up to 85 percent was attained, while shot grading accuracy was up to 75 percent.