CVMar 25, 2014

Capturing and Recognizing Objects Appearance Employing Eigenspace

arXiv:1403.6260v1
Originality Synthesis-oriented
AI Analysis

This addresses the challenge of object recognition in cluttered environments for robotics applications, but it appears incremental as it builds on existing eigenspace methods.

The paper tackles the problem of automatically capturing and recognizing object appearances from complex scenes using robotic vision, by constructing eigenspaces that include both occluded and non-occluded views, with experimental results demonstrating the approach's effectiveness.

This paper presents a method of capturing objects appearances from its environment and it also describes how to recognize unknown appearances creating an eigenspace. This representation and recognition can be done automatically taking objects various appearances by using robotic vision from a defined environment. This technique also allows extracting objects from some sort of complicated scenes. In this case, some of object appearances are taken with defined occlusions and eigenspaces are created by accepting both of non-occluded and occluded appearances together. Eigenspace is constructed successfully every times when a new object appears, and various appearances accumulated gradually. A sequence of appearances is generated from its accumulated shapes, which is used for recognition of the unknown objects appearances. Various objects environments are shown in the experiment to capture objects appearances and experimental results show effectiveness of the proposed approach.

Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes