AI-Driven Detection of Moisture Issues in Buildings Using Convolutional Neural Networks

Abstract

This study presents the development and application of a Convolutional Neural Network (CNN) model for the detection and classification of moisture-related issues in buildings, specifically focusing on leakage, seepage, and dampness. The study highlights the challenges faced in data acquisition, including the similarities in visual characteristics of different moisture issues and the limitations posed by varying lighting conditions and image quality. A high-resolution mobile camera was employed to capture images, which were then processed using a CNN model. The results demonstrate the model’s accuracy and reliability, with an overall classification accuracy of 94%.