Smart City Gnosys
Smart city article details
| Title | Monocular Vision Based Approach For Occlusion Detection And Handling: A Way Forward For Advanced Driver Assistance Systems |
|---|---|
| ID_Doc | 37916 |
| Authors | Upadhyaya V.; Tiwari N. |
| Year | 2024 |
| Published | International Journal of Intelligent Transportation Systems Research, 22, 1 |
| DOI | http://dx.doi.org/10.1007/s13177-024-00389-1 |
| Abstract | With the advent of the concept of IOT for smart cities and autonomous vehicles, Intelligent Transport system has sprouted as a new research area. The major thrust areas are security, surveillance and coordination in transport infrastructure and Advanced Driver Assistance Systems in smart Vehicles. It requires real time information collection of road and traffic monitoring. Though LIDARS, high end cameras and various sensors are deployed for this purpose but due to limitations in their sensing capabilities they still require human intervention for effective surveillance. In accordance with the type of sensors and their location of deployment, various approaches for detection and classification of vehicles have been developed. All these vision- based traffic monitoring approaches basically rely on object detection which is a classical area of research in the field of computer vision. Various novel applications viz. navigation assistance for high end vehicles and autonomous vehicles, traffic monitoring and surveillance find their roots in this area. This paper aims at presenting a cost effective and simple monocular vision-based approach vis a vis RADAR and LIDAR approach for Occlusion Detection and Handling under different scenarios. The proposed approach is based on the three-dimensional representation of location of the object being tracked in order to mitigate the problem of occlusion effectively. The proposed methodology gives 97.6% of precision viz a viz the contemporary object detection techniques. © The Author(s), under exclusive licence to Intelligent Transportation Systems Japan 2024. |
| Author Keywords | Occlusion Detection; Occlusion handling; Tracking; Vision based object detection |
