Computer Vision based Fitness Assessment using Mobile Devices

The identification and development of athletic talent in various regions remain a significant challenge due to limited access to professional coaching and standardized evaluation system. To address this problem, the paper presents an AI powered Fitness Assessment platform designed to democratize sports talent identification. The suggested system uses realtime computer vision techniques to assess physical fitness metrics using standardized scoring, including pushups, squats, vertical jumps, planks, endurance runs, and shuttle runs. Benchmarking based on gender and age guarantees equity. It integrates pose estimation through media pipe based skeletal tracking, motion analysis, repetition count-ing via custom state machine logic, biomechanical validation algorithms for form assessment. A more scalable, transparent, and inclusive framework for national athletic talent discovery is provided by automating manual and resource-intensive fitness testing procedures. This is especially important in developing nations where traditional systems struggle with affordability, accessibility, and scalability.