Content Details | NUML Online Research Repository

Content Details

Back to Department Listing

Title Crack Detection in Solar Panel Using B-Net Deep Learning Model

Author(s) Bilal Buta

Abstract Renewable energy is seen as an alternative to fossil fuel consumption to reduce environmental pollution. Solar energy is considered the most potential renewable energy source since it is economical and energy-efficient. Photovoltaic panels are susceptible to physical damage which can significantly decrease efficiency and lead to expensive repairs. Conventional methods are laborious and inclined to human error, emphasizing the need for automated systems. In this work, the B-Net model is developed to detect cracks in solar panels. It is based on a convolutional neural network architecture to enhance accuracy and effectiveness in identifying cracks under various lighting and weather conditions. An inclusive dataset containing cracked and non-cracked images of solar panels is employed to enable the B-Net model to learn differentiating features effectively. Findings indicate that the model attains high accuracy and precision in defect detection, better than traditional techniques. Moreover, the B-Net model’s performance metrics, such as accuracy, precision, recall, loss, and F1-score, are analyzed to determine its effectiveness. This work contributes to the maintenance of solar systems and prepares the path for further enhancement in automated assessment technologies through deep learning models. The implications of this work extend beyond photovoltaic panel maintenance, offering comprehensive applicability to other fields requiring image-based crack detection.

Type Thesis/Dissertation MS

Faculty Engineering and Computer Science

Department Mathematics

Language English

Publication Date 2025-02-24

Subject Mathematics

Publisher

Contributor(s)

Format

Identifier

Source

Relation

Coverage

Rights