Detecting of photovoltaic installations in RGB aerial imaging: a comparative study

Abstract

ABSTRACT:In this work, we compare four different approaches for detecting photovoltaic installations from RGB aerial images. Our client, an electricity grid administrator, wants to hunt down fraud with unregistered illegal solar panel installations by detecting installations in aerial imagery and checking these against their database of registered installations. The detection of solar panels in these RGB images is a difficult task. Reasons are the relatively low resolution (at 25 cm/pixel an individual solar panel only measures about 9 × 7 pixels), the undiscriminating colour properties of the object (due to in-class variance and specular effects) and the apparent shape variability (rotation and skew due to the different roofs slant angles). Therefore, straightforward object segmentation techniques do not yield a satisfying solution, as proven in this paper. We compared four state-of-the-art object detection approaches for this task. First we experimented with a machine learning object detection technique based on pixel-based support vector machine classification. Secondly we developed an approach using MSER based colour segmentation and shape analysis. Finally a dual approach based on object categorization using the boosted cascade classifier technique of Viola & Jones and the aggregate channel features technique of Dollár et al., is introduced, learning a combination of colour and gradient feature based classifiers from a given training set. We successfully evaluate these four different approaches on a fully labelled test set of a 8000 × 8000 pixel, 4 square km zone containing 315 solar panel installations with in total more than 10.000 individual panels.

Detection And Area Estimation For Photovoltaic Panels In Urban Hyperspectral Remote Sensing Data By An Original Nmf-Based Unmixing Method

“…However, this approach is costly and time consuming, which leads to the use of other faster and less expensive methods such as the integration of remote sensing data. Recently, some approaches, using remote sensing data, were proposed for the automatic detection of photovoltaic installations and their localization [1], [2]. These methods use high spatial resolution airborne/spaceborne RGB images, therefore with a limited number of spectral bands.…”

Automatic Boundary Extraction for Photovoltaic Plants Using the Deep Learning U-Net Model

“…The image segmentation techniques with TIP were developed to identify objects such as the area of PV Plants out of an orthophoto [10,34,35]. Later, the Machine Learning (ML) and Deep Learning (DL) image segmentation techniques, also known as semantic segmentation, were proposed [36,37].…”

Partial Linear NMF-Based Unmixing Methods for Detection and Area Estimation of Photovoltaic Panels in Urban Hyperspectral Remote Sensing Data

“…The use of remote sensing data is an interesting approach to automatically detecting photovoltaic installations. In this context, some remote sensing-based approaches have recently been proposed [1,4]. These approaches exploit very high-spatial-resolution airborne/spaceborne RGB images with a small number of spectral bands.…”