Research on fingertip positioning and human-computer interaction technology based on stereo vision

Abstract

Human-computer interaction system based on computer vision is an important research direction in the field of human-computer interaction, which has broad application prospects in natural human-computer interaction, sign language recognition, virtual reality, smart home, somatosensory games and other fields. In this paper, an interactive visual perception method is proposed, which uses human experience to guide the computer to quickly build the visual perception model. By analyzing the difference between the predicted image and the actual image read by the camera, the area with large reflectivity change on the projection screen is found as the user area. This method can find the position of real human hand correctly even if the projection image contains human hand. Three-dimensional coordinates of fingertips are obtained by stereo matching principle, and Kalman filtering tracking algorithm is used to smooth the trajectory of fingertips and narrow the detection range of the next frame.