"Sensor fusion" Science-Research, February 2022 -- summary from Arxiv

Approximating and comprehending the surroundings of the vehicle exactly creates the critical and basic step for the autonomous vehicle. Based upon recent research, 3D things discovery structures performing item detection and localization on LiDAR data and sensor fusion methods reveal considerable improvement in their performance. In this paper, we present a parallel style for a sensor fusion detection system that combines a video camera and 1D light discovery and varying sensors for object detection. Using a spatio-temporal placement and a policy of sensor fusion, we finished the advancement of a fusion discovery system with high integrity at ranges of up to 20 m. Test results showed that the suggested system attains a high level of accuracy for pedestrian or things detection before a vehicle, and has high robustness to unique environments.