A fast LiDAR-based features extraction/tracking using hough transforms and fuzzy C-means clustering for LiDAR-aided Multisensor Navigation Systems
This paper proposes a fast feature extraction/tracking methodology for LiDAR-Aided multisensor integrated navigation systems. Hough Transform is applied on the LiDAR range/bearing information in 2D space to detect lines. To filter out noisy observations and outliers and focus only on strong line patterns, a fuzzy C-mean clustering algorithm is utilized. By tracking extracted lines features, the relative 2D orientation/translation motions are estimated. The proposed methodology was applied on an unmanned ground vehicle (UGV) to estimate its 2D relative orientation/translational motion. The estimated LiDAR-based relative orientation/translational changes are fused with Inertial/Odometer measurements by an Extended Kalman Filter (EKF). The integrated solution was compared with Inertial/Odometer standalone navigation output and results showed significant improved accuracy when LiDAR updates are applied.
|Conference||27th International Technical Meeting of the Satellite Division of the Institute of Navigation, ION GNSS 2014|
Nematallah, H. (H.), Lui, S. (S.), Atia, M, Givigi, S. (S.), & Noureldin, A. (A.). (2014). A fast LiDAR-based features extraction/tracking using hough transforms and fuzzy C-means clustering for LiDAR-aided Multisensor Navigation Systems. In 27th International Technical Meeting of the Satellite Division of the Institute of Navigation, ION GNSS 2014 (pp. 3184–3193).