基于LeNet网络的图像识别应用研究

Application of Image Recognition Based on LeNet Network

卷积神经网络(CNN)在图像识别方面起着无法替代的作用,其经典网络模型包括LeNet、AlexNet、VggNet等。笔者以MINIST数据集作为训练样本,在DIGITS上使用LeNet网络进行模型训练,根据数据集特征进行适用情境预测,采用相应图像测试模型准确率,验证适用情境预测的正确性,并结合测试结果对模型适用情境进行分析。该结论可以作为后续根据使用情境训练专用网络模型的参考。

Convolutional neural network(CNN)plays an irreplaceable role in image recognition.Its classic network models include lenet,alexnet,vggnet,etc.In this paper,Minist data set is used as the training sample,and lenet network is used to train the model on digits.According to the characteristics of the data set,the applicable situation prediction is carried out.The accuracy of the corresponding image test model is used to verify the correctness of the applicable situation prediction,and the applicable situation of the model is analyzed based on the test results.This conclusion can be used as a reference for training the private network model based on the use of context.