Abstract
The evolution of motion capture technology from marker-based to marker-less systems is a promising field, emphasizing the critical role of combining electromyography (EMG), joint moments, and ground reaction forces (GRF) in advancing biomechanical analysis. This review examines the integration of EMG, joint moments, and GRF in marker-less video-based motion capture systems, focusing on current approaches, challenges, and future research directions. This paper recognizes the significant challenges of integrating the aforementioned modalities, which include problems of acquiring and synchronizing data and the issue of validating results. Particular challenges in accuracy, reliability, calibration, and environmental influences are also pointed out, together with the issue of the standard protocols of multimodal data fusion. Using a comparative analysis of significant case studies, the review examines existing methodologies' successes and weaknesses and established best practices. New emerging themes of machine learning techniques, real-time analysis, and advancements in sensing technologies are also addressed to improve data fusion. By highlighting both the limitations and potential advancements, this review provides essential insights and recommendations for future research to optimize marker-less motion capture systems for comprehensive biomechanical assessments.