作者: Dirk Holz , Ruwen Schnabel , David Droeschel , Jörg Stückler , Sven Behnke
DOI: 10.1007/978-3-642-20217-9_11
关键词:
摘要: For planning grasps and other object manipulation actions in complex environments, 3D semantic information becomes crucial. This paper focuses on the application of recent Time-of-Flight (ToF) cameras context scene analysis. being able to acquire from ToF camera data, we a) pre-process data including outlier removal, filtering phase unwrapping for correcting erroneous distance measurements, b) apply a randomized algorithm detecting shapes such as planes, spheres, cylinders. We present experimental results that show robustness against noise outliers underlying RANSAC paradigm allows segmenting classifying objects captured natural mobile setups.