Parallel landing sites detection using LiDAR data on manycore systems
Helicopters are widely used in emergency situations where knowing if a geographical location is adequate for landing is a critical issue, and it is far from being a straightforward task. In this work, we present an ongoing project to use LiDAR point clouds to detect and classify landing sites in real time. Landing sites are detected in parallel on manycore systems. Load balancing was identified as the main cause of poor performance. We present results for a set of LiDAR point clouds and balancing strategies for three different multi and manycore systems. The load balancing techniques applied
increase performance up to 3 times from the unbalanced case.
keywords: LiDAR, landing, load balancing, Xeon Phi
Publication: Congress
1624015041889
June 18, 2021
/research/publications/parallel-landing-sites-detection-using-lidar-data-on-manycore-systems
Helicopters are widely used in emergency situations where knowing if a geographical location is adequate for landing is a critical issue, and it is far from being a straightforward task. In this work, we present an ongoing project to use LiDAR point clouds to detect and classify landing sites in real time. Landing sites are detected in parallel on manycore systems. Load balancing was identified as the main cause of poor performance. We present results for a set of LiDAR point clouds and balancing strategies for three different multi and manycore systems. The load balancing techniques applied
increase performance up to 3 times from the unbalanced case. - Lorenzo, Oscar G.; Martínez, Jorge; Vilariño, David L.; Pena, Tomás F.; Cabaleiro, José C.; Rivera, Francisco F.
publications_en