Advancing Utility Pole and Sign Detection Through Deep Learning

2026-08-04

AI Summary

This study addresses the challenge of detecting wooden utility poles and their attached warning signs in ground-level images, where blurred appearance and weak visual features hinder accurate detection and tilt-angle estimation. To this end, the authors propose a unified framework based on an enhanced Detection Transformer (DETR) that simultaneously performs pole detection, instance segmentation, geometric tilt-angle estimation, and sign classification. This work represents the first application of DETR to this complex real-world scenario and introduces the first publicly available benchmark dataset for ground-view pole-sign analysis, comprising 4,570 annotated images. Experimental results demonstrate high detection accuracy, with mean average precision (mAP) of 90.43% for poles and 88.26% for signs; furthermore, among 1,433 test poles, the tilt angle is accurately estimated for 1,367, achieving a mean absolute error of only 1.01 degrees.

Abstract

Utility poles are an essential part of the infrastructure used to support power distribution systems and other critical public services. Their regular inspection is crucial to ensure the stability and safety of the electrical grid. A deep learning framework is presented for the automated detection, segmentation and lean angle estimation of wooden utility poles, and classification of attached electrical warning signs, using ground-level imagery. The system is trained on a custom dataset of 4,570 annotated images extracted from Google Street View, featuring challenging real-world scenes with visually ambiguous wooden poles lacking distinctive features. The proposed model is based on the Detection Transformer (DETR), suitably modified and trained on the custom dataset. The model outperforms standard object detectors (RetinaNet, Faster R-CNN, YOLOv3-Tiny), achieving a mean average precision of 90.43% for pole detection and 88.26% for sign detection. Extending this model with a segmentation head enables per-instance mask generation, which is then used to estimate pole lean angle. The model accurately estimates lean for 1,367 out of 1,433 test-set poles, with a mean absolute error of 1.01 degrees. Moreover, the custom dataset created in this work is also made publicly available to be used as a benchmark.

Problem

• utility pole detection
• sign classification
• lean angle estimation
• infrastructure inspection
• ground-level imagery

Innovation

• Detection Transformer
• instance segmentation
• lean angle estimation
• utility pole detection
• custom dataset