Estimating the ball possession ratio is a big challenge due to the overlapping of objects, the difficulty in locating the ball, and the fast movement of players and referees on the field. To address these limitations, this paper proposes an integrated Homographic Perspective Transform (HPT) method to transform the space from the image to the football field, combined with a hybrid YOLO detector improved by the Weighted Intersect Fusion (WIF) technique, to enhance the ability to accurately detect players and balls. Objects are tracked across multiple frames using the Geometric Midpoint (GM) method, which helps to accurately estimate the ball possession ratio of each team. At the same time, a distance threshold-based technique is applied to determine the player holding the ball. The experimental results of the proposed method show that it is relatively superior to the existing methods.