Surface Estimation of a Pedestrian Walk for Outdoor Use of Power Wheelchair
Independent thesis Advanced level (degree of Master (Two Years)), 20 credits / 30 HE creditsStudent thesis
In mobile robotics surface estimation and object recognition plays a vital role for navigationand control. This research presents a normal vector estimation method of a surfaceusing Delaunay tessellation. The method is an expansion of previously developed ContinuousNearest Neighbor algorithm, underlining the trade-o between ltering and quality ofinput data. First the 3D data points are segmented through a threshold process. Second,normal vectors are found using an averaging method of centroids on Delaunay tessellation.Eect of dierent parameters inuencing the performance of the algorithm is provided.Moreover, two similarity measures of vector angle and Euclidean distance are considered forsurface estimation of a pedestrian walk. Two dierent scenarios are set to test and analyzethe robustness of the algorithm. Apart from this another data clustering method usingMahalanobis distance is used so that the surfaces can be distinguished.
Place, publisher, year, edition, pages
Electrical Engineering, Electronic Engineering, Information Engineering
IdentifiersURN: urn:nbn:se:umu:diva-58390OAI: oai:DiVA.org:umu-58390DiVA: diva2:548376
Master's Programme in Robotics and Control