Solar panel current classification model

How can Pul-RF classify solar panels?

In the second stage, when using PUL-RF, the classifier could learn the spectral characteristics of PV solar panels across multiple background types. Additionally, during the second stage of sampling, unlabelled samples were collected from various surface types, thus enabling the differentiation of solar panels from other surface types (Fig. 6).

What are the spatial distribution characteristics of PV solar panels in 2022?

From the spatial distribution characteristics of PV solar panels in 2022 (Fig. 7a), global PV is concentrated in the middle and low latitudes, and there is little PV distribution in high latitudes. Besides, PV solar panels are mostly distributed in densely populated areas of the world except Africa.

How can we identify PV Panels globally?

We developed a new method to identify PV panels globally, producing an annual 20-meter resolution dataset for 2019–2022. This dataset offers unprecedented detail and accuracy for future research and policy-making. A two-stage PV classification framework was built using U-Net and positive unlabelled learning with random forest (PUL-RF).

Can a large set of PV solar panels be identified as positive samples?

Due to the prior participation in training U-Net with PV solar panel labels covering various background types such as cultivated land, forest land, artificial surfaces, deserts, mountains, and water bodies, in the first stage, a relatively rich set of PV solar panels could be identified as positive samples for the second stage classification.