Full-Text Search:
Home|Journal Papers|About CNKI|User Service|FAQ|Contact Us|中文
《Journal of Jiamusi University(Natural Science Edition)》 2019-06
Add to Favorite Get Latest Update

Research on Image Classification Method Based on Improved Multi-channel Convolutional Neural Network Model

ZHOU Yan-ting;College of Mathematics and Big Data, Anhui University of Science and Technology;  
In order to fully extract image feature information and reduce the overfitting, this paper proposes an improved multi-channel convolutional neural network model. First, the improved model extracts image feature information by using three convolution channels. Each channel sets different sizes of convolution kernel, and reduces model parameters by stacking small convolution kernel instead of large convolution kernel, and then process feature information by feature fusion and batch normalization technology. Finally, the feature information is input to the softmax classifier to classify. The improved model is compared with single-channel model, multi-channel model and traditional image classification model for CIFAR-10 dataset. The experimental results show that the improved model can effectively extract image feature information, reduce the over-fitting, and improve the classification accuracy of the model.
【Fund】: 国家自然科学基金(11601007)
【CateGory Index】: TP391.41;TP183
Download(CAJ format) Download(PDF format)
CAJViewer7.0 supports all the CNKI file formats; AdobeReader only supports the PDF format.
©2006 Tsinghua Tongfang Knowledge Network Technology Co., Ltd.(Beijing)(TTKN) All rights reserved