In this paper we present work in progress on the development of a low-cost autonomous robotic platform that integrates multiple state-of-the-art techniques in RGB-D perception to form a system capable of completing a real-world task in an entirely autonomous fashion. The task we set out to complete is determining the location of a preselected object within the physical world. This experiment requires a robotic framework with a number of capabilities including autonomous exploration, dense real-time localisation and mapping, object detection, path planning and motion control.
|Title of host publication||RSS Workshop on RGB-D: Advanced Reasoning with Depth Cameras|
|Number of pages||4|
|Publication status||Published - 1 Jul 2014|