In the large-screen interactive system with lidar sensor, due to the low accuracy of the lidar and the instability of the users’ gestures, the system’s recognition and tracking of gesture coordinates cannot be well obtained. Aiming at solving the problems of swaying and drifting gestures of the traditional filtering algorithm with a lidar sensor, this paper proposes a contactless interaction control technology based on switching filtering algorithm, which can realize non-contact high-precision multi-point interaction. The proposed algorithm first recognizes and extracts users’ gestures, and then the gestures are mapped to the screen position. Also, the mouse operation is simulated to realize operations such as selecting, sliding, and zooming in and out. Besides, the algorithm can effectively solve jitter and drift problems caused by scanning defects of radar and instability of the user gesture operations. Experimental results show that by applying the switching filtering algorithm to the contactless human-computer interaction system, the interactive trajectory becomes smoother and more stable compared with the traditional filtering algorithms. The proposed algorithm exhibits excellent accuracy and real-time performance, supporting efficient interaction with multiple people.
