DocumentCode :
1718524
Title :
Object recognition by dynamic link matching with multiple blob formation
Author :
Umeki, Hideo ; Mizutani, Hiroyuki
Author_Institution :
Syst. & Software Eng. Lab., Toshiba Corp., Kawasaki, Japan
fYear :
1996
Firstpage :
237
Lastpage :
245
Abstract :
We present a one-to-many object matching system based on neural dynamic link architecture. When an input image containing multiple objects is given, if some of them are similar to a stored model, the system can establish geometric transformation-invariant mappings between the model and the corresponding object regions in the input image. This can be achieved by extending the fast dynamic link matching (FDLM) algorithm to allow multiple blob formation. Numerical simulations of neural layer dynamics indicate that multiple blobs can be developed where the layer input is sufficiently strong against the background level. To extract matched regions from the input layer, we consider each neural layer as a graph and introduce another neural system based on local edge mappings. This system can roughly detect the matched regions with neighborhood-preserving mappings without global cost functions
Keywords :
computational geometry; image matching; neural net architecture; object recognition; dynamic link matching; fast dynamic link matching algorithm; geometric transformation-invariant mappings; image recognition; matched region detection; multiple blob formation; neighborhood-preserving mappings; neural dynamic link architecture; object recognition; one-to-many object matching system; Computer architecture; Face recognition; Laboratories; Neurons; Numerical simulation; Object recognition; Pattern matching; Research and development; Software engineering; Solid modeling;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Neural Networks for Identification, Control, Robotics, and Signal/Image Processing, 1996. Proceedings., International Workshop on
Conference_Location :
Venice
Print_ISBN :
0-8186-7456-3
Type :
conf
DOI :
10.1109/NICRSP.1996.542765
Filename :
542765
Link To Document :
بازگشت