Abstract <p>This article describes the development of an intelligent decision support system for creating individual training plans for professional athletes using computer vision technologies. The emphasis is on the need to take into account the individual physiological characteristics of athletes, differences in the structure of injury causes by sport and age group, as well as the importance of adapting training loads to minimize the risk of injury. The importance of synthesizing the ontological approach, artificial intelligence, and computer vision for creating a system that allows for the formation of a universal environment that facilitates the process of developing and adapting ontologies is emphasized. A system is developed to analyze athletes’ physical data obtained using computer vision and create personalized training plans based on it, which in turn helps reduce injuries and improve the effectiveness of the training process. The study also discusses the use of meta-associative graphs to create digital twins of athletes, which is a progressive approach in the context of training systems. The results confirm the potential of using the developed system to optimize the training of athletes in various disciplines.</p>