本文已被：浏览 1909次 下载 3753次
Received:June 21, 2011 Revised:July 17, 2011
Received:June 21, 2011 Revised:July 17, 2011
中文摘要: 给出了一种基于LBP 算子和BP 神经网络的医学图像分类方法。该方法采用方向纹理谱描述符来描述图像的纹理特征,并以此特征作为网络的输入对BP 神经网络进行训练,最后采用已训练的BP 网络进行医学图像的分类。通过实验表明,该方法可以很好地对医学图像进行分类识别。
Abstract:A method of medical images classification based on LBP operator and BP neural network is proposed in this paper. The method uses direction texture spectrum descriptor to describe the texture features, making them as the samples input of neural network for training. Finally, the trained network is used to classify the medical images. Experimental results show it is effecitve for medical images classification.
keywords: medical images classification local binary patterns direction texture spectrum BP neural network
文章编号： 中图分类号： 文献标志码：
基金项目:教育部科学技术研究重点项目(210128);河南理工大学博士基金(B050901)
引用文本：
孙君顶,李琳.基于BP 神经网络的医学图像分类.计算机系统应用,2012,21(3):160-162,212
SUN Jun-Ding,LI Lin.Classification of Medical Image Based on BP Neural Network.COMPUTER SYSTEMS APPLICATIONS,2012,21(3):160-162,212
孙君顶,李琳.基于BP 神经网络的医学图像分类.计算机系统应用,2012,21(3):160-162,212
SUN Jun-Ding,LI Lin.Classification of Medical Image Based on BP Neural Network.COMPUTER SYSTEMS APPLICATIONS,2012,21(3):160-162,212
