A motion accuracy evaluator based on body parts movement by MapReduce video processing

This paper presents a healthcare application that tracks body parts movement in video recording persons who exercise, analyzes the motor performance (e.g., motion speed and space), and evaluates fitness status (e.g., motion accuracy and abnormality). We design a MapReduce video processing for collecting the data of body parts movement from a large number of video files and successfully shorten the execution time. Analyzing the video files, the application correctly detects persons who have difficulties in their motion speed and space.