Photovoltaic panel roof identification

What is the quality of PV panel identification?

In summary, the quality of the PV panel identification is very high (high OA). The lower PA and UA is mainly due to the low spatial resolution of the HySpex data as well as the geometric displacement between the validation and HySpex data. 5.3. Future directions

Can solar panels be detected on a roof?

Although various authors have already shown that CV can be effectively used to detect solar panels in satellite images on a large scale [ 5, 6, 7 ], only a few studies have explored detecting the whole range of superstructures on a roof. Mainzer et al. [ 25] used conventional CV techniques to extract roof segments and roof superstructures.

Which Visualization Library is used for rooftop photovoltaics?

The library for visualization is matplotlib. The project target is to segment in aerial images of Switzerland (Geneva) the area available for the installation of rooftop photovoltaics (PV) panels, namely the area we have on roofs after excluding chimneys, windows, existing PV installations and other so-called ‘superstructures’.

Can roof superstructure detection be used for PV potential analysis?

First, we present the Roof Information Dataset (RID) for roof segments and roof superstructures. Second, we examine the annotation quality of roof superstructures and its influence on neural networks’ training and prediction. Third, we assess the viability and benefit of using CV-based roof superstructure detection for PV potential analysis.

What is a typical color feature of PV panels?

This phenomenon indicates that the typical color feature of PV panels is a pattern of monotonous color with contrasting colors. The critical point for the gap of different groups of color features is the brightness instead of color pattern, as shown in Fig. 12 (a).

Can U-nets be used to segment roof-top PV panels in satellite images?

In 2020, Zhuang et al. proposed a cross-learning driven U-Net (CrossNets) method to segment roof-top PV panels in satellite images. However, the above studies focused on using the universal machine learning frameworks such as CNN, U-Net, DeepLabv3 and etc., lacking analyzing the characteristics of PV image data and improving the models.