Most of the patients suffering from shoulder movement impairment have some problems in daily life. Consequently, shoulder movement rehabilitation plays an important role for them and needs to be evaluated in continued treatments of the shoulder. In addition to the cost of treatment, the travel expenses to visit a hospital are also increasing. This paper proposes a shoulder angle measurement (SAM) system using computer vision with a web camera. The system detects arm in an image and calculates active shoulder angle movement. A correlation analysis of the measured shoulder angle between the SAM system and two devices, namely ATAN Scale (Adapted Thai Arthrometric Navigator Scale) and a goniometer, was performed. The Pearson correlations between the SAM system and the two devices were close to one. The maximum full-scale errors were 5.03 and 3.69 on comparing to the ATAN Scale and the goniometer, respectively.
