COMBINED CLOSE RANGE PHOTOGRAMMETRY AND REMOTE SENSING FOR PHOTOVOLTAIC PARKS EFFICIENCY ANALYSIS

Abstract

The paper presents an innovative approach of combining the methods of close-range photogrammetry and remote sensing in the analysis of the efficiency of photovoltaic parks. A UAV was used to capture the site and close-range photogrammetry technology was used to create orthoimages in the thermal and visible zone. Orthoimages obtained from close-range photogrammetry imagery were combined, and the resulting multichannel image was analysed with remote sensing methods. Spatially overlapping RGB and thermal orthoimages are collocated and a multispectral image was created. Cluster analysis of the multispectral image was performed. Using remote sensing methods, a four-channel multichannel image was created and classified through supervised classification approach with predefined masks. The training data for the classifier represents the types of defects found on the photovoltaic panels. The obtained results were analysed and an assessment of the classification was made.