Robust on-line model-based object detection from range images

Abstract

A mobile robot that accomplishes high level tasks needs to be able to classify the objects in the environment and to determine their location. In this paper, we address the problem of online object detection in 3D laser range data. The object classes are represented by 3D point-clouds that can be obtained from a set of range scans. Our method relies on the extraction of point features from range images that are computed from the point-clouds. Compared to techniques that directly operate on a full 3D representation of the environment, our approach requires less computation time while retaining the robustness of full 3D matching. Experiments demonstrate that the proposed approach is even able to deal with partially occluded scenes and to fulfill the runtime requirements of online applications. © 2009 IEEE

Robust place recognition for 3D range data based on point features

“…His approach extracts Spin Images [11] from 3D scans and uses them to match each scan against a database. In our previous work [15] we demonstrated, that the features we use in this paper are more descriptive than Spin Images in the context of object recognition. Huber reported 1.5 s as the time requirement to match one scan against another.…”

Detecting Objects in Scene Point Cloud: A Combinational Approach

“…Another category deals with indoor scenes. [7] used a model-based method to detect chairs and tables in the office. [8] aims at classifying objects such as office chair.…”

Hierarchical object geometric categorization and appearance classification for mobile manipulation

“…While their presented recall curves outperform others, the number of objects is relatively low and household objects are less distinct. In [11], the authors investigate the extraction of GOODSAC point features and object recognition from range images, that are in turn computed from point cloud datasets. These object models are, as in our case, created from real 3D data but processed using the work in [9].…”