Classifying point cloud datasets to identify distribution wires is essential for detecting vegetation encroachment near power lines. These workflows help prevent fires and power outages but are often manual, repetitive, and labour-intensive.

This model is specifically designed to extract distribution wires at the street level. Its performance is more consistent for street-level distribution wires, while predictions for high-tension transmission wires vary with geographic changes. Consequently, the model shows a lower recall rate for high-tension transmission wires compared to the higher recall observed for low-lying street wires and poles.