This Project work proposes a real-time approach for multi-vehicle detection and tracking. It involves detecting and continuously monitoring the status of vehicles from an aerial perspective, encompassing factors such as vehicle speed, vehicle classification, counting in areas under CCTV surveillance. Multicamera tracking is a computer vision task that involves analyzing videos to identify and track objects belonging to one or more categories. New technologies like object detection and tracking have been developed to utilize automated camera surveillance and generate data that can inform decision-making processes. For example, if the police need to locate a suspect's car in an area covered by hundreds of cameras, monitoring all the surveillance videos from each camera and reviewing thousands of hours of footage would be challenging. However, with computer vision technology, we can detect the location of vehicles and filter large amounts of data based on their shape, type, and appearance. In terms of vehicle identification, we implemented a spatial attention mechanism that relies on the background model. Our system's effectiveness was evaluated through the camera vehicle identification task.
Multi Camera Vehicle Tracking Using OpenCV & Deep Learning
24.11.2023
986804 byte
Aufsatz (Konferenz)
Elektronische Ressource
Englisch
Face detection and tracking: Using OpenCV
IEEE | 2017
|Springer Verlag | 2022
|Drowsiness Detection System Using OpenCV
IEEE | 2025
|Underneath Vehicle Inspection Using Fuzzy Logic, Subsumption and OpenCV Library
BASE | 2015
|Automatic Vehicle Number Plate Recognition System (AVNPR) Using OpenCV Python
Springer Verlag | 2021
|