Defect object detection algorithm for electroluminescence image defects of photovoltaic modules based on deep learning

Abstract

Visual inspection of photovoltaic modules using electroluminescence (EL) images is a common method of quality inspection. Because human inspection requires a lot of time, object detection algorithm to replace human inspection is a popular research direction in recent years. To solve the problem of low accuracy and slow speed in EL image detection, we propose a YOLO‐based object detection algorithm YOLO‐PV, which achieves 94.55% of AP (average precision) on the photovoltaic module EL image data set, and the interference speed exceeds 35 fps. The improvement of speed and accuracy benefits from the targeted design of the network architecture according to the characteristics of EL image. First, we weaken the backbone's ability to extract deep‐level information so that it can focus on extracting the low‐level defect information. Second, the PAN network is used for feature fusion in the Neck part. But, only the single‐size feature map output is retained, which significantly reduces the amount of calculation. Also, we analyze the impact of data enhancement methods on model overfitting and performance. Finally, we give effective data enhancement methods. The results show that the object detection algorithm in this paper can meet the requirements for high‐precision and real‐time processing on the PV module production line.

PV Cell Defect Detection Based on Improved YOLOv7 Modeling

“…The utilization of solar energy as a green, pollution-free renewable energy has seen a dramatic rise in recent years [1]. Currently, photovoltaic cells are the primary means of converting solar energy into electrical energy, and are the fundamental components of power generation systems.…”

An intelligent defect detection method in solar cells via IDATLGAN

“…The former method relies on domain experts or field technicians to identify defects through professional knowledge and practical experience by naked eyes, which not only increases the work intensity but also increases the time cost greatly. However, the latter mainly extracts the features from EL images automatically by the deep learning-based models [5] to detect defects in an end-to-end manner [6]. Although the deep learning-based models is superior to the labor-intensive methods to some extent and successfully applied in fault diagnosis [7], spectrum sensing [8] and other fields, Models based on deep learning are susceptible to the problems aroused by lack of sufficient labeled data and class imbalance in practical applications [9].…”

Identification and localization of Photovoltaic Defect from UAV Thermal Infrared Data Based on Photogrammetry and Deep Learning

“…There are two types of fault detection for solar panels: detection based on electrical signals (current and voltage) and detection based on images (vision) [3,4]. Thermal infrared images with a resolution of up to 0.02 m have proven to be sufficient for detecting defect types in practical applications, and thermal infrared photos with a resolution of 0.02 m are more efficient at the photographic stage than millimeter-scale RGB images.…”