Integrating Spatiotemporal and Visual Features for Enhanced Object Re-identification in Multi-camera Networks for Intelligent Transportation Systems
摘要
In this paper, we propose a novel approach to enhancing object re-identification in multi-camera networks within Intelligent Transportation Systems (ITS). The proposed method integrates trajectory-based prediction and visual feature analysis, and considers the installation conditions of cameras to overcome performance variations. Through experiments with two cameras on an actual road, we show the feasibility of the proposed method.