The study aims to verify the accuracy of a developed hardware-software system for analyzing human posture using a digital video camera. The objective of the work was to compare the results obtained with the developed system to the data from the Vicon Nexus motion capture system. The research included the development of a methodology for processing video images to determine posture parameters, the creation of software for automatic video analysis and metric calculation, and the conduct of an experimental study involving 14 participants (7 men and 7 women). Simultaneous recording was performed using a digital video camera (4K resolution, 30 fps) and the Vicon Nexus system with infrared cameras and reflective markers. Data analysis employed computer vision techniques, including the use of the pre-trained neural model SAM for image segmentation and the computation of angular posture characteristics. Statistical analysis demonstrated a high degree of agreement between the two systems (cross-correlation coefficient $r = 0.81$), with a result discrepancy of 4–5%. Key factors affecting accuracy include technical limitations of the video camera, errors in mathematical models, and optical system distortion. The obtained results confirm the potential of using the developed system for posture assessment in clinical settings, opening up prospects for its application in sports medicine, rehabilitation, and biomechanics.