A preprocessing method and classification of 2D objects using neural networks

被引：0

作者：

Topalova, I

Tzokev, A.

Ilchev, S.

机构：

来源：

ANNALS OF DAAAM FOR 2007 & PROCEEDINGS OF THE 18TH INTERNATIONAL DAAAM SYMPOSIUM: INTELLIGENT MANUFACTURING & AUTOMATION: FOCUS ON CREATIVITY, RESPONSIBILITY, AND ETHICS OF ENGINEERS | 2007年

关键词：

digital data processing; 2D object classification; optimization; neural network;

D O I：

暂无

中图分类号：

TP [自动化技术、计算机技术];

学科分类号：

0812 ;

摘要：

The paper suggests a pre-processing method and classification of 2D objects using neural network aiming high recognition rate. The proposed research is based on development of a method for extracting object parameters and choosing an appropriate pre-processing to represent an input vector to a neural network (NN) classification system. The input vector itself and the structure of the NN are optimized in the search for a minimal number of neurons, fast classification time and high classification rate.

引用

页码：767 / 768

页数：2