Paper Info
Title
A Neural Network for Symmetry-Based Object Detection and Tracking
Abstract
A Discrete Neural Network (DNN) is presented determining a measure of similarity between two images. The network provides a correspondence map linking points of both images depending on the characteristics and arrangement of their neighborhood. The DNN is applied to the detection of bilateral symmetries in images taken from scenes of motor ways. A set of hypothetical symmetry axis is evaluated by similarity measurement between the left and right image halves, respectively.
Year
DOI
Venue
1996
10.1007/978-3-642-80294-2_5
DAGM-Symposium
Keywords
Field
DocType
symmetry-based object detection,neural network
Object detection,Pattern recognition,Computer science,Time delay neural network,Left and right,Artificial intelligence,Artificial neural network,Homogeneous space
Conference
ISBN
Citations 
PageRank 
3-540-61585-7
1
0.38
References 
Authors
4
2
Name
Order
Citations
PageRank
thomas kalinke113426.81
W von Seelen2503140.13