Full-Text Search:
Home|Journal Papers|About CNKI|User Service|FAQ|Contact Us|中文
《Computer and Modernization》 2018-07
Add to Favorite Get Latest Update

Traffic Sign Recognition Based on Convolutional Neural Network

CHEN Bai-li;LIN Nan;College of Mathematics and Informatics,South China Agricultural University;  
With the development of science and technology,the recognition of traffic signs for unpiloted vehicle is becoming more and more practical. And the most popular method to recognize the traffic sign is to use the convolutional neural network. The experiment is made based on this method. After the collection of actual traffic signs image data,we use the histogram-equalization and other methods to achieve image preprocessing and constitute the training set and test set. Then we build the convolution neural network of Le Net5 by the training set,and use the test set to test the accuracy of the model. In the first experiment,the recognition accuracy of the model is higher than other traditional identification methods,but only 87. 5%. After the adjustment of the neural network's batch size,the convolution kernel,the number of training and other aspects,the recognition accuracy is increased to 93. 2%.
【CateGory Index】: TP183;TP391.41
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