International Journal of Scientific Research and Engineering Development

Tracking and recovering misplaced items in colleges, offices, and public spaces is a common problem that manual methods handle poorly. This paper describes a web-based lost-and-found system that uses machine learning to match reported lost items with found items submitted by other users. Images are processed through a pre-trained MobileNetV2 model to generate visual feature vectors, while item descriptions are converted into numerical representations using TF-IDF. A combined similarity score ranks potential matches and notifies users when a likely match is found. A simple claim process allows users to confirm and collect their items. The system is built using Flask and SQLite and is designed to be easy to deploy within an institution.

Mrs.Divyashree K, Mrs.Smitha B N, Jeevan Paragi, Kiran D, Manish Harasur, Rohith C A, "Web-Based Lost and Found System with Object Matching Using Machine Learning" International Journal of Scientific Research and Engineering Development, V9(4): Page(834-836) July-August 2026. ISSN: 2581-7175. www.ijsred.com. Published by Scientific and Academic Research Publishing.