A Study of Intelligent Rehabilitation Robot Imitation of Human Behavior Based on Kinect

In recent years, the number of stroke patients has gradually increased, and the serious imbalance between the number of rehabilitation physicians and patients has led to a large number of patients unable to perform the necessary rehabilitation training, which affects the rehabilitation effect of patients. The emergence of medical rehabilitation robots on the basis of existing robots. The integration of the teaching strategy can effectively reduce the workload of rehabilitation physicians, while the existing teaching rehabilitation robots use less visual fusion, and the manipulation method is complicated and cumbersome. In response to the above situation, this paper proposes a Kinect camera-based rehabilitation robot schematic behavior recognition method to realize the task of using machine vision to perceive the trajectory of the rehabilitation physician’s schematic teaching task to the patient and map it onto the rehabilitation robot.