Title :
Extracting essential local object characteristics for 3D object categorization
Author :
Madry, Marianna ; Afkham, Heydar Maboudi ; Ek, Carl Henrik ; Carlsson, Stefan ; Kragic, Danica
Author_Institution :
Centre for Autonomous Syst. & the Comput. Vision & Active Perception Lab., KTH R. Inst. of Technol., Stockholm, Sweden
Abstract :
Most object classes share a considerable amount of local appearance and often only a small number of features are discriminative. The traditional approach to represent an object is based on a summarization of the local characteristics by counting the number of feature occurrences. In this paper we propose the use of a recently developed technique for summarizations that, rather than looking into the quantity of features, encodes their quality to learn a description of an object. Our approach is based on extracting and aggregating only the essential characteristics of an object class for a task. We show how the proposed method significantly improves on previous work in 3D object categorization. We discuss the benefits of the method in other scenarios such as robot grasping. We provide extensive quantitative and qualitative experiments comparing our approach to the state of the art to justify the described approach.
Keywords :
feature extraction; image representation; object detection; 3D object categorization; feature occurrences; local object characteristics extraction; object classes; object representation; robot grasping; summarization technique; Animals; Databases; Feature extraction; Histograms; Robustness; System-on-chip; Three-dimensional displays;
Conference_Titel :
Intelligent Robots and Systems (IROS), 2013 IEEE/RSJ International Conference on
Conference_Location :
Tokyo
DOI :
10.1109/IROS.2013.6696670