In this paper we present an effective and robust system to classify fruits under varying pose and lighting conditions tailored for an object recognition system on a mobile platform. Therefore, we present results on the effectiveness of our underlying segmentation method using RGB as well as depth cues for the specific technical setup of our robot. A combination of RGB low-level visual feature descriptors and 3D geometric properties is used to retrieve complementary object information for the classification task. The unified approach is validated using two multi-class RGB-D fruit categorization datasets. Experimental results compare different feature sets and classification methods and highlight the effectiveness of the proposed features using a Random Forest classifier.
[pdf]
@inproceedings{jiang_ROBIO2013,
author = {Jiang, Lixing and Koch, Artur and Scherer, Sebastian A. and Zell,
Andreas},
title = {Multi-class Fruit Classification using {RGB-D} Data for Indoor Robots},
booktitle = {IEEE International Conference on Robotics and Biomimetics (ROBIO)},
year = {2013},
address = {Shenzhen, China},
month = {December},
abstract = {In this paper we present an effective and robust system to classify
fruits under varying pose and lighting conditions tailored for an
object recognition system on a mobile platform. Therefore, we present
results on the effectiveness of our underlying segmentation method
using RGB as well as depth cues for the specific technical setup
of our robot. A combination of RGB low-level visual feature descriptors
and 3D geometric properties is used to retrieve complementary object
information for the classification task. The unified approach is
validated using two multi-class RGB-D fruit categorization datasets.
Experimental results compare different feature sets and classification
methods and highlight the effectiveness of the proposed features
using a Random Forest classifier.},
url = {http://www.cogsys.cs.uni-tuebingen.de/publikationen/2013/jiang_robio2013.pdf}
}