Lost Sharing Web Service Based on Image Classification Model

Abstract

Losing personal belongings is a common occurrence, yet the recovery rate of lost items remains alarmingly low. For instance, in South Korea, the recovery rate of lost phones is merely 3.2%, despite the efforts of mobile carriers and the police. Existing lost and found platforms, such as community-based applications, lack the necessary features for effective recovery, resulting in limited success. To address this issue, this paper proposes a lost and found sharing web service that utilizes image classification models to improve the identification and recovery of lost items. By enabling users to upload images and detailed information about lost or found belongings, this service enhances communication between users and increases the likelihood of successful recovery. The proposed system combines a user-friendly community forum with advanced image matching technology to better connect lost items with their rightful owners. The paper presents a detailed system design and discusses the potential impact of the proposed service on improving lost item recovery rates.