Utility Poles Identification using Google Street View Images

Accurate detection of utility poles is essential for effective infrastructure management in power systems, but traditional methods often fall short in delivering the needed precision. Previous versions of You Only Look Once (YOLO) showed promise but lacked the accuracy required for reliable utility pole identification. To address these shortcomings, an enhanced YOLOv9 model was developed to improve detection performance in street view images. By incorporating advanced feature extraction techniques, the YOLOv9 model achieves better detection accuracy. Tested on Google Street View images, the model reached a 95% detection accuracy while reducing the False Positive Rate to $\mathbf{9. 0 1 \%}$. These results demonstrate the model’s ability to accurately identify utility poles with fewer false alarms, offering a reliable tool for infrastructure supervision. The enhanced YOLOv9 marks a significant improvement, making maintenance and management more efficient.