Efficient Region Segmentation of PV Module in Infrared Imagery using Segnet. Issue 1 (June 2021)

• Record Type: Journal Article
• Authors: Xie, Ying; Shen, Yu; Zhang, Kanjian; Zhang, Jinxia
• Is Part Of: IOP conference series. Volume 793:Issue 1(2021)
• Publication Date: 2021-06
• Subjects: Earth sciences -- Periodicals; Environmental sciences -- Congresses; Environmental sciences -- Periodicals; 550.5
• DOI: 10.1088/1755-1315/793/1/012018
• Languages: English
• ISSNs: 1755-1307

Abstract

Abstract: As renewable energy, solar energy resources are a major focus. The flaw detection of the PV production system is an important guarantee for the stable operation of the system. Hotspot detection is a key step. It is very important to extract the efficient region in the infrared image of the photovoltaic module in advance to improve the hot spot detection precision. In this paper, we propose an effective region segmentation method for infrared image of photovoltaic module based on SegNet, which greatly improves the calculation efficiency and detection accuracy. We use mask processing to hide the irrelevant background area in the original image and label the image data with labelme software. We trained and validated the model using infrared images of photovoltaic modules captured by the portable infrared imager provided by the electric company, and we assessed the model. This paper is the first attempt to use deep learning technology to solve the engineering problem of effective region segmentation of photovoltaic module infrared image. The experimental results show that the segmentation effect of our proposed methodology is remarkable in practical technical applications.