Identification and Tracking of Main Material of Transmission Tower Based on Monocular RGB Photography

This paper investigates the recognition and tracking of transmission tower main materials in single RGB photography environment. Considering the photographic noise in complex environments, this paper models the photographic image using line features, analyzes the 3D-2D camera photographic model of transmission tower main material features, the spatial motion model of rigid bodies and the motion model of transmission tower main material line features in photographic images. A recognition and tracking method for transmission tower main materials is designed, which integrates Kalman filtering and probability modeling. Experimental results show that the proposed method can effectively track transmission tower main materials and exhibits good robustness in complex environments. At the same time, the method proposed in this paper also has high speed, accuracy and stability, and can meet the needs of tracking and recognition of transmission tower main materials in power transmission and distribution systems.