Aerial photography of the principle of solar panels

How do we detect solar panel locations using aerial imagery?

We use deep learning methods for automated detection of solar panel locations and their surface area using aerial imagery. The framework, which consists of a two-branch model using an image classifier in tandem with a semantic segmentation model, is trained on our created dataset of satellite images.

Do solar panels have object detection models?

Reports of solar panel installations have been supplemented with object detection models developed and used on openly available aerial imagery, a type of imagery collected by aircraft or drones and limited by cost, extent, and geographic location.

Can satellite imagery be used to identify solar PV systems?

One possible solution to this problem is to identify existing solar PV generation systems using overhead satellite and aerial imagery. While there have been early promising attempts in this direction, there are nevertheless many important research challenges that remain to be addressed.

Can deep convolutional networks be used for aerial imagery?

The poor generalization of deep convolutional networks to aerial imagery from new geographic locations: an empirical study with solar array detection. In 2017 IEEE Applied Imagery Pattern Recognition Workshop (AIPR), 1–8 (IEEE, 2017).

Can satellite imagery improve solar panel detection accuracy?

We address these limitations by providing a solar panel dataset derived from 31 cm resolution satellite imagery to support rapid and accurate detection at regional and international scales. We also include complementary satellite imagery at 15.5 cm resolution with the aim of further improving solar panel detection accuracy.

Are annotated solar panels available in native resolution and HD satellite imagery?

To the best knowledge of the authors, there are no publicly available datasets including annotated solar panels in native resolution and HD satellite imagery. The process for creating the paired native resolution and HD image tiles and associated labels. Both sets of components contain three image tiles and 2,542 annotated solar panels.