Vortex-Induced Vibration Identification of Transmission Tower Components with the Improved Feature Matching Method Based on the Camera Video

Accurately identifying the vortex-induced vibration (VIV) responses of transmission tower members is crucial for assessing their dynamic characteristics under wind loads, preventing fatigue damage, and enabling intelligent structural maintenance. Traditional manual inspection and contact-sensor-based methods struggle to rapidly and accurately capture such sudden and transient vibration events due to limitations in deployment efficiency, sensor coverage, and cost. To address this, this paper proposes a fast, non-contact identification method for the VIV responses of transmission tower members with a consumer camera. First, to tackle the challenge of unstable tracking caused by sparse surface features at long distances, a virtual extension of the region of interest (VEROI) method is established. This method intelligently incorporates contextual features from both the target member and its adjacent areas, effectively increasing the number of feature points available for stable tracking while suppressing background interference. Second, an improved marginalizing sample consensus (IMAGSAC) feature matching algorithm is proposed to enhance the accuracy and robustness of displacement extraction. By incorporating motion consistency constraints and an efficient mismatch filtering strategy, the reliability of the extracted vibration displacement time history is significantly improved. The adaptability and accuracy of the proposed framework are systematically verified through comparative laboratory shaking table experiments and field tests on an actual tower under wind excitation. Finally, a sudden VIV event of a member was successfully identified. This study provides an efficient and practical technical solution for the intelligent inspection and online monitoring of abnormal vibrations in transmission towers, contributing to the digital and predictive transformation of power infrastructure maintenance.