Bowler's Real-Time Posture Assessment Using Integrated Moving Body and Frame Difference with Kinect and Image Processing Algorithms

• Published in: International Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment and Management (Online) pp. 1 - 6
• Main Authors: Balbin, Jessie R., Valiente, Flordeliza L., Cruz, Christian Rae S. Dela, Gregorio, John Robin G., Tinaya, Patrick Carlo J., Villarba, Joel Andrew P.
• Format: Conference Proceeding
• Language: English
• Published: IEEE 19.11.2023
• Subjects: angle comparison; bowling; four-step approach; image processing; Imaging; Kinect; Legged locomotion; Libraries; posture; Process control; PyKinect; python; Real-time systems; slope; Statistical analysis; Training
• ISSN: 2770-0682
• DOI: 10.1109/HNICEM60674.2023.10589029

This paper discusses a system that detects bowling posture and corrects it by image analysis using frame difference and image processing algorithms. Using a Kinect as the main sensor, we were able to create a system that uses skeletal imaging to obtain coordinate data of the body which is converted to angle form and compared with a reference with a set tolerance. Python was used as the main programming language because of the easy integration of the Py Kinect library that is responsible for the manifestation of the skeletal frame on the subject. The four-step approach was used as the reference for the correct bowling posture and was assessed by a professional bowler. The results taken were also shown to the coach to ensure accuracy. Correlation Analysis was used as the statistical tool to prove the similarity of the data taken from the program and the evaluation of the professional coach.