This research addresses the challenge of creating an advanced tennis training system that replicates a player’s unique play style [1]. Leveraging machine learning and computer vision, the system analyzes video footage to extract shot characteristics such as speed, spin, and trajectory, which are then used to control a tennis ball machine for realistic practice [2]. The methodology incorporates a mobile application for video analysis, a neural network-based detection model for object recognition, and hardware integration for precise shot replication. Experiments demonstrated the system’s high accuracy in daylight conditions but highlighted challenges under low light and with hardware limitations in spin control [3]. Comparative analysis with other methodologies shows that this approach uniquely combines video analysis and physical replication, offering a realistic and personalized training experience. By integrating advanced technologies with tennis training, this system bridges the gap between analysis and application, providing athletes with an innovative tool to refine their skills and enhance performance.