3-D Shape Context The 3-D shape context (SC) descriptor [6] is an extension of the 2-D SC descriptor [7] into 3-D. The descriptor for a given key point is computed utilizing a 3-D spherical grid, centered on the key point, and superimposed over all points in the point cloud. Within the support structure, a set of sectors is defined by means of subdivisions along the azimuth, elevation, and radial dimensions, equally spaced for the first two dimensions and logarithmically spaced for the radial dimension. The final descriptor is built from the subdivision of the grid structure and is represented in bins of a 3-D histogram. Each bin of the histogram isassociated with a value and computed as the weighted sum of the number of points falling in the corresponding grid sector. Additionally, the computed weight is inversely proportionaltothelocalpointdensityandthebinvolume. Thenorthpoleofthe3-Dgridisalignedwiththenormal computedonthekeypointbeingdescribed,whilenoprincipal direction isdefined overthenormal plane. Thisremainingdegreeoffreedomisdealtwithbytakingintoaccountas manyorientationsofthegridasthenumberofazimuthsubdivisions, leading to several local reference frames and, in turn, to multiple descriptions for the same key point to deal with rotations and translations of the point cloud. Algorithm2 shows how to compute the 3-DSCdescriptor using PCL. Again, observe the similarity between this algorithm and the one presented in Algorithm 1, requiring the user to setanadditionaltwoparameters.