mfranzon/solar-panels-yolov8s-seg

Aerial Solar Panels (YOLOv8s-seg)

Instance segmentation of individual PV arrays from drone and nadir aerial imagery, with tight masks on the panels themselves. Part of the Model Garden collection.

Class

solar-panel

Provenance

Weights derive from `finloop/yolov8s-seg-solar-panels`, re-hosted here for the Model Garden showcase. On a Wikimedia solar-farm test set it locked panels at conf 0.90 to 0.97 on 6 of 8 shots, missing only very distant obliques.

Training

• —Base: YOLOv8s-seg
• —Task: instance segmentation

Usage

from ultralytics import YOLO

model = YOLO("finloop_seg.pt")
results = model("aerial.jpg")
results[0].show()