A New Approach for Hot Spot Solar Cell Detection based on Multi-level Otsu Algorithm

Abstract

Hotspot cells can harm the photovoltaic modules because the cells spend or use up energy instead of generating energy. Therefore, the image processing method is applied to allow the hot spot detection automatically. This paper evaluates Multi-level Otsu based approach for image processing to segment and detect hot spot solar photovoltaic cell of photovoltaic module using thermal images. The methodology was completed by performing image segmentation, image binarization, and noise removal with morphological operation. This proposed method was tested by using ten thermal images and evaluated by measuring the accuracy of each image with different threshold level. The highest accuracy achieved 97.22% from thermal image DJI_0898, DJI_0911 and DJI_0918.