Dean Sacoransky is a researcher and MASc candidate at Queen’s University in Kingston, Ontario, Canada, with a focus on intelligent sensors, autonomous systems, and artificial intelligence. He is co-supervised by Dr. Keyvan Hashtrudi-Zaad and Dr. Joshua Marshall. He earned his BASc degree in computer engineering from Queen’s University in 2022. As a graduate research assistant and member of the Queen’s University AutoDrive Challenge II Team, Dean is actively engaged in research and development related to autonomous vehicles. Dean has previously worked on projects related to computer vision, sensor calibration, and applied artificial intelligence. Above all, Dean is dedicated to leveraging technology to improve human safety and enhance quality of life.
Dean’s research is focused on all-weather autonomous vehicle localization. Dean is developing a multi-object tracking system, which estimates the relative position of reflective lane markers through automotive radar, Kalman filter state stimation, and artificial intelligence.
Dean has gained hands on experience interfacing with various sensors and robots (camera, LiDAR, radar, IMU, wheel encoders, Clearpath Husky A200). Dean has used the Robot Operating System (ROS) to interface with hardware and deploy real-time perception and navigation algorithms.
Autonomous vehicle localization system using radar and reflective lane markers
This project focusses on the formulation and experimental validation of an autonomous ground vehicle localization method that applies a single modality, millimeter wave radar perception system for the detection of roadside retro-reflectors. There are factors that make radar-based perception systems challenging, such as the sparsity of radar point clouds and the impact of motion-induced noise on radar sensor performance. To overcome these challenges, we propose a radar feature extraction scheme that uses Kalman filter state estimation and DBSCAN clustering to segment retro-reflector points from noise points, and thus provide the autonomous vehicle with a predicted path for the road ahead.
Publications and Videos
This work will be presented at ICRA 2023 in London, United Kingdom.
D. Sacoransky, K. Hashtrudi-Zaad, and J. A. Marshall. Towards unsupervised filtering of millimetre-wave radar returns for autonomous vehicle road following. To appear in Proceedings of the 2023 IEEE International Conference on Robotics & Automation (ICRA), London, UK, May-June 2023. [Preprint PDF]
Through the Ingenuity Labs Research Institute at Queen’s University, we have access to a Clearpath Husky A200 Robot, a Continental ARS-408 radar sensor, Brewers’ 12-inch aluminium retro-reflectors, and a Vicon motion capture system for ground truth validation. We conducted experiments to test our system on curved and straight-shaped paths and varying vehicle ego speeds.
Contact and Media
- email: firstname.lastname@example.org