Hotspot Detection in Photovoltaic Module using Otsu Thresholding Method

Abstract

The growth of enthusiasm in developing power from sustainable resources has promoted to rise in the setting up of photovoltaic system. However, after the installations, it is still very low awareness of checking the PV module condition. Early hotspot detection is necessary to keep the PV working. Thermal imaging is still considered as common technique for diagnosing hotspot. This study proposes to develop a method to recognize hotspot in photovoltaic module thermal images using image processing technique. There are three essential process consisting of pre-processing, segmentation with Otsu thresholding method and identification. The data used contains 6 cropped images. The average accuracy results achieve 92.16%. This value shows that the proposed method is suitable to contribute the development of hotspot detection in PV monitoring system.

Access to Document

Other files and links

• Link to publication in Scopus

Fingerprint