The system locates human bone data based on a three-point camera. The source of bone data includes two parts. One part is the color data collected by the color camera, and the other part is the depth data collected by the infrared camera. The color camera obtains the target RGB image information, the color histogram is obtained through the incremental histogram algorithm, and then the weight is calculated by the particle filter algorithm. The depth camera obtains the target depth information through the optical coding technology, and uses the color data and the depth data to merge and calculate the weight. Way to achieve tracking of the target object. The core technology of the human gait bone data location extraction evaluation system based on three-point camera is depth measurement technology, which is a kind of optical coding technology. The light source illuminates the space to be measured, and the objects in the space reflect light from different positions. Then, the three-dimensional space is calibrated once by the light source, and finally the specific structure of the entire space is obtained.  Camera-based somatosensory interaction technology can directly use body movements to interact with digital devices and virtual environments, which solves the problem of using or wearing complex force tactile interaction and motion capture devices in the existing virtual interaction process, and can track and track in real time. It captures the movement position of the limbs during exercise, and combines human-computer interaction technology to perform sports rehabilitation training, providing patients with a novel rehabilitation environment, helping to improve the patient's self-care ability and speed up the patient's recovery. The system is applied to rehabilitation centers and can initially help bone rehabilitation personnel to provide rehabilitation evaluation reports.