Research on Image Edge Detection Algorithm Based on Eigenvector and Improved BP Neural Network

DOI

10.2991/ammee-17.2017.128How to use a DOI?

Keywords

Edge detection, eigenvector, BP neural network, improved algorithm, filter out noise.

Abstract

The purpose of edge detection is to distinguish different regions of the image, usually the edge information is determined by the gray-scale change between regions. In the image with noise, the traditional edge detection method is easy to determine the noise as the edge point. While adding filtering measures can reduce noise, it also reduces edge information. In this paper, an edge detection method based on eigenvector and improved BP neural network is proposed. The eigenvector of the pixel is composed of differential and median value of the grayscale. Select the sample image to extract the eigenvector of each pixel and enter the BP network for training. Through the error between output value and the tutor signal to adjust network parameters. The improved BP network with learning rate self-regulation and momentum factor can improve the network performance [1]. After the training is completed, use the Cameraman image for edge detection and compared with the traditional detection methods. Experiments show that the method can remove the noise while preserving the edge.

Copyright

© 2017, the Authors. Published by Atlantis Press.

Open Access

This is an open access article distributed under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).

Download article (PDF)

TY  - CONF
AU  - Mingyang Yu
AU  - Xiaoyu Huang
PY  - 2017/06
DA  - 2017/06
TI  - Research on Image Edge Detection Algorithm Based on Eigenvector and Improved BP Neural Network
BT  - Proceedings of the Advances in Materials, Machinery, Electrical Engineering (AMMEE 2017)
PB  - Atlantis Press
SP  - 667
EP  - 673
SN  - 2352-5401
UR  - https://doi.org/10.2991/ammee-17.2017.128
DO  - 10.2991/ammee-17.2017.128
ID  - Yu2017/06
ER  -