By Topic

Neural network-based regions detection

Sign In

Cookies must be enabled to login.After enabling cookies , please use refresh or reload or ctrl+f5 on the browser for the login options.

Formats Non-Member Member
$33 $13
Learn how you can qualify for the best price for this item!
Become an IEEE Member or Subscribe to
IEEE Xplore for exclusive pricing!
close button

puzzle piece

IEEE membership options for an individual and IEEE Xplore subscriptions for an organization offer the most affordable access to essential journal articles, conference papers, standards, eBooks, and eLearning courses.

Learn more about:

IEEE membership

IEEE Xplore subscriptions

1 Author(s)
T. Ohyama ; Wakayama Shin-Ai Women's Junior Coll., Japan

Plural regions extraction takes its important place in image processing, pattern recognition and computer vision research. Its objective is to locate prominent regions in a noisy image, and to extract the components for feature recognition. The region detection can be achieved by some image processing techniques, feature detection operators and neural network approaches. If we need plural regions in an image, such as characters in a map, these conventional methods are usually unsuccessful because they cannot distinguish the desirable regions from noisy background. Neocognitron and filtering neural network are solutions of these problems. They detect a most prominent region in an image, but they are difficulty to detect plural regions in an image simultaneously. It is common practice to use evidence from biological and psychological experiments to develop the computational model for the neural network. This paper introduces a method of plural regions extraction in a noisy image using the three-layered neural network. This model has the mutual overlapped receptive field of adjoining hidden layer units. It achieves the plural-regions detection and its generalization ability through the backpropagation learning. Finally, we present an example of the experimental results to show the performance of this method

Published in:

Neural Networks, 1995. Proceedings., IEEE International Conference on  (Volume:3 )

Date of Conference:

Nov/Dec 1995