Year: 2025
Type: article
Abstract: This paper presents HomeSafe, an AI-driven system for automated housing inspection using Machine Learning-based anomaly detection. The system detects structural cracks from images and estimates their severity, providing detailed reports for decision-making. A MobileNetV2-based CNN model is fine-tuned for crack detection, achieving high accuracy on a benchmark dataset. The system further estimates crack width and dimen sions using image processing techniques. A web-based interface (Streamlit) is developed for easy image uploads and report generation. The project aims to assist engineers, homeowners, and real estate professionals in conducting automated building inspections. Key Words: Crack Detection, Deep Learning, Anomaly De tection, CNN, MobileNetV2, Image Processing, Streamlit, Hous ing Inspection.
DOI: 10.55041/ijsrem44600
Authors: Dr. Amrapali Chavan, Soham Sakunde
Institution: AISSMS Institute of Information Technology
Language: English
Subfield: Civil and Structural Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: bronze
OpenAlex ID: W4409443791