Plane-based object categorisation using relational learning
Author(s)
Farid, Reza
Sammut, Claude
Griffith University Author(s)
Year published
2014
Metadata
Show full item recordAbstract
We use Inductive Logic Programming (ILP) to learn classifiers for generic object recognition from point clouds, as generated by 3D cameras, such as the Kinect. Each point cloud is segmented into planar surfaces. Each subset of planes that represents an object is labelled and predicates describing those planes and their relationships are used for learning. Our claim is that a relational description for classes of 3D objects can be built for robust object categorisation in real robotic application. To test the hypothesis, labelled sets of planes from 3D point clouds gathered during the RoboCup Rescue Robot competition are used ...
View more >We use Inductive Logic Programming (ILP) to learn classifiers for generic object recognition from point clouds, as generated by 3D cameras, such as the Kinect. Each point cloud is segmented into planar surfaces. Each subset of planes that represents an object is labelled and predicates describing those planes and their relationships are used for learning. Our claim is that a relational description for classes of 3D objects can be built for robust object categorisation in real robotic application. To test the hypothesis, labelled sets of planes from 3D point clouds gathered during the RoboCup Rescue Robot competition are used as positive and negative examples for an ILP system. The robustness of the results is evaluated by 10-fold cross validation. In addition, common household objects that have curved surfaces are used for evaluation and comparison against a well-known non-relational classifier. The results show that ILP can be successfully applied to recognise objects encountered by a robot especially in an urban search and rescue environment.
View less >
View more >We use Inductive Logic Programming (ILP) to learn classifiers for generic object recognition from point clouds, as generated by 3D cameras, such as the Kinect. Each point cloud is segmented into planar surfaces. Each subset of planes that represents an object is labelled and predicates describing those planes and their relationships are used for learning. Our claim is that a relational description for classes of 3D objects can be built for robust object categorisation in real robotic application. To test the hypothesis, labelled sets of planes from 3D point clouds gathered during the RoboCup Rescue Robot competition are used as positive and negative examples for an ILP system. The robustness of the results is evaluated by 10-fold cross validation. In addition, common household objects that have curved surfaces are used for evaluation and comparison against a well-known non-relational classifier. The results show that ILP can be successfully applied to recognise objects encountered by a robot especially in an urban search and rescue environment.
View less >
Journal Title
Machine Learning
Volume
94
Issue
1
Subject
Computer Vision
Artificial Intelligence and Image Processing
Information Systems
Cognitive Sciences