FURI | Fall 2021

Real-Time Object Detection Using Image and Lidar Data

Data icon, disabled. Four grey bars arranged like a vertical bar chart.

Lidar and camera sensor data are utilized, along with an object detection model, to recognize vehicles and their distances from the sensor system. This research project would help drivers by serving as a tool to help recognize potential dangers on the road. It also has the potential to be integrated into the car’s system to help with safety measures, such as collision avoidance. This system is implemented on an NVIDIA Jetson AGX Xavier to take advantage of its embedded GPU so that the system can be run as real-time as possible.