A Novel Algorithm for Detection of Soccer Ball and Player

Abstract

In this work, we propose an efficient, mere algorithm for detection of a soccer ball and players. In long shots, existing algorithms may not be able to detect the ball when it merged with the lines in the field. So, we introduce a method that can separate lines from the ball and segment the ball competently. Primarily, the ground and the edges of the original image are detected. Then, Hough transform line detection algorithm is applied to identify lines. After line detection, the lines are eliminated from the original image so that only detected players and the ball remain in the scene. The experimental results show how durable that our algorithm can detect ball when the ball attached to the line and also can detect players under occluded situations.