DEVELOPMENT OF SOFTWARE FOR RECOGNITION OF DANCE ELEMENTS

. The development of software for the analysis of dance elements is an interdisciplinary task combining computer vision and machine learning methods. The main purpose of such systems is to automate the process of motion analysis, which minimizes the subjectivity of assessment and improves diagnostic accuracy. The article discusses key aspects of the development, including algorithms for constructing coordinate matrices based on key body points, data normalization methods, and motion trajectory analysis. Special attention is paid to identifying the strengths and weaknesses of the system, such as the high accuracy of determining key points and dependence on the quality of input data. Methods of eliminating disadvantages are proposed, including the use of multi-chamber systems, data filtering using a Kalman filter, and the use of adaptive neural networks to account for individual user characteristics. Recommendations for setting up the equipment, including the optimal location of cameras and lighting, are considered.