A Novel Approach for Vehicle Theft Detection using Image Processing

Abstract

The aim of the project is to identify the stolen vehicles by comparing the images taken by the CCTV cameras in a given location using image processing and neural networks. As video analytic technologies are developing, object detection and classification of video data can be performed by machines like CCTV cameras and computers. The automatic object detection and classification can be applied to identify same vehicle from multiple video clips taken at different time or by different cameras. The accuracy of video object classification can be enhanced by using other information. The proposed project involves technologies such as image processing and neural networks to integrate the independently working cameras and to make them work together so that the identification of a stolen vehicle is possible automatically. The concerned authorities are alerted on the identification of the stolen vehicle automatically.