Obstacle Identification from 3D Data for AGVs in a Warehouse Environment
|Title||Obstacle Identification from 3D Data for AGVs in a Warehouse Environment|
|Summary||Obstacle Identification from 3D Data for AGVs in a Warehouse Environment|
|Keywords||3D point cloud, time of flight camera, obstacle detection, segmentation, object recognition, mobile robot|
|TimeFrame||Start: February 2014, End: June 2014|
|References|| Zhang, Hao, et al. "SVM-KNN: Discriminative nearest neighbor classification for visual category recognition." Computer Vision and Pattern Recognition, 2006 IEEE Computer Society Conference on. Vol. 2. IEEE, 2006.
Golovinskiy, Aleksey, Vladimir G. Kim, and Thomas Funkhouser. "Shape-based recognition of 3D point clouds in urban environments." Computer Vision, 2009 IEEE 12th International Conference on. IEEE, 2009.
Rusu, Radu Bogdan, and Steve Cousins. "3d is here: Point cloud library (pcl)." Robotics and Automation (ICRA), 2011 IEEE International Conference on. IEEE, 2011.
Nüchter, Andreas, and Joachim Hertzberg. "Towards semantic maps for mobile robots." Robotics and Autonomous Systems 56.11 (2008): 915-926.
Lai, Kevin, and Dieter Fox. "Object recognition in 3D point clouds using web data and domain adaptation." The International Journal of Robotics Research 29.8 (2010): 1019-1037.
Brostow, Gabriel J., et al. "Segmentation and recognition using structure from motion point clouds." Computer Vision–ECCV 2008. Springer Berlin Heidelberg, 2008. 44-57.
Rusu, Radu Bogdan, et al. "Fast 3d recognition and pose using the viewpoint feature histogram." Intelligent Robots and Systems (IROS), 2010 IEEE/RSJ International Conference on. IEEE, 2010.
Drost, Bertram, et al. "Model globally, match locally: Efficient and robust 3D object recognition." Computer Vision and Pattern Recognition (CVPR), 2010 IEEE Conference on. IEEE, 2010.
|Prerequisites||Image analysis, machine learning, programming skill, ROS and PLC|
|Supervisor||Björn Åstrand, Saeed Gholami Shahbandi|
A very essential element to achieve the proper solution of the intelligent warehouses, is AGVs with smart behavior. One criteria of a smart behavior is the way vehicles handle obstacle encountering. The goal of this project is to use a 3D sensor (Fotonic P70, a time of flight camera) to detect and identify the obstacles appearing in the path of AGV (lift-trucks) in warehouses. Research Question: while the current solution to obstacle avoidance for lift-trucks in the work environment involves a set of 2D range sensors and obstacle detection, desired result of this project is to develop a method for obstacle identification by mean of a 3D sensors, in order to increase “situation awareness” of AGVs and behave more intelligently.
Work package 1: 3D point cloud manipulation (system setup) Work package 2: object detection (segmentation) Work package 3: identity recognition of obstacles (classification) Work package 4: estimating the motion of obstacles from a sequence of frames (bonus part)
Deliverable: an implementation and demonstration of a developed method for obstacle identification.