With the continuous maturity of deep learning technology, its role in stage character design is becoming more and more important. The article proposes a deep neural network-based multi-view human tracking method for stage environment and designs the stage lighting control system. And on this basis, it designs an automatic stage lighting tracking system, including control and management module, actor identification and localization module, data processing module, and speed interpolation module, which facilitates the control of lighting on the stage. Finally, a series of experimental tests are conducted to verify the effectiveness of the method of this paper. The experimental results on simulated and real labeled datasets show that the binary cyclic code of this paper’s method can still achieve more than 91% recognition accuracy under 60% occlusion rate, which has a very good anti-occlusion performance. Without affecting the stage performance and viewing experience, this paper’s method solves the tracking instability problem caused by stage darkness and actor’s apparent similarity, which is highly feasible and has a wide range of application prospects in the stage performance industry.
