Title :
Towards automatically learning an implicit model from 2D-images based on a local similarity analysis of contours
Author :
Pechtel, D. ; Kuhnert, K.-D.
Author_Institution :
Inst. for Realtime Data Process., Siegen Univ., Germany
Abstract :
The article deals with enabling an intelligent system to autonomously learn an implicit model of its environment. An unsupervised learning method is presented which learns the topological connections of different object views. Moreover, the method is able to distinguish between different objects. Based on a systematic local analysis of the objects´ contours, the method unites learning a topology (i.e. navigation) and object recognition into one framework
Keywords :
image matching; image recognition; intelligent control; mobile robots; object recognition; path planning; robot vision; topology; unsupervised learning; 2D images; automatic learning; autonomous learning; contour analysis; environment model; implicit model; intelligent mobile robots; intelligent system; local similarity analysis; navigation learning; object recognition; object views; systematic local analysis; topological connections; topology learning; unsupervised learning; Data processing; Ear; Image sequences; Intelligent robots; Learning systems; Navigation; Organisms; Robot vision systems; Solid modeling; Topology;
Conference_Titel :
Intelligent Robots and Systems, 2000. (IROS 2000). Proceedings. 2000 IEEE/RSJ International Conference on
Conference_Location :
Takamatsu
Print_ISBN :
0-7803-6348-5
DOI :
10.1109/IROS.2000.894668
