Assorted classes to facilitate robotic vision, typically in the post processing of RGBD point clouds. Standard transforms such as Canny edge detector, 2D Hough transform, and 3d Kernel Hough Transform, along with new paradigms of stereo disparity matching and robotic navigation.