Akademik Veri Yönetim Sistemi
2024 11th International Conference on Electrical and Electronics Engineering (ICEEE), Muğla, Türkiye, 22 Nisan 2024, ss.377-381
Due to the limitations of low-cost cameras in object tracking, it may be challenging to obtain reliable data from recordings captured by such cameras, particularly when dealing with moving objects. In this work, a novel technique was established to precisely identify and monitor the motions of the tennis ball and ground contact locations using low-resolution and low-frame-rate video recordings. The research paper presents a methodology that effectively identifies and monitors the rapid motions of a tennis ball, regardless of fluctuations in illumination and low frames per second values. By combining techniques for image enhancement, motion estimation, and adaptive filtering, the developed algorithm attempts to overcome the shortcomings of existing systems. Even under demanding conditions, the results of this study demonstrate that the algorithm achieves a high rate of success in detecting objects during sports analyses captured with low-resolution cameras. In summary, this study presents a novel approach to enhance precision and availability in the domains of sports science and performance analysis, particularly in settings with limited resources.