摘要
随着大数据时代的到来,数据挖掘、图像处理等已经成为了一个热门研究方向。本文的研究目的是自动识别猫狗类型,采用的是基于数据挖掘的猫狗自动识别技术。本文将位于全方位下拍摄的具有外貌复杂的猫狗图像运用卷积神经网络训练。本实验挑选前沿的深度学习框架pytorch以及计算能力强大的GPU,使用深度神经网络VGG16,分别对猫狗图像进行网络训练与测试。实验显示使用VGG16网络模型进行识别的准确率非常高,在猫狗类型识别问题上具有突出优势。
With the advent of the era of big data,data mining and image processing has become the hot research direction.The purpose of this paper is to automatically identify the cat and dog types using the cat and dog automatic recognition technology based on deep learning.In this paper,convolution neural network is used to train the cat and dog images with complex appearance.In this experiment,the researchers select the advanced deep learning framework pytorch and the powerful GPU,and use the deep neural network VGG16 to train and test the cat and dog images.Experiments show that the accuracy of VGG16 network model is very high,and it has outstanding advantages in cat and dog type recognition.
作者
王彬
张正平
贾明俊
陆安江
卢学敏
WANG Bin;ZHANG Zhengping;JIA Mingjun;LU Anjiang;LU Xuemin(College of Big Data and Information Engineering,Guizhou University,Guiyang 550025,China)
出处
《智能计算机与应用》
2021年第7期162-165,共4页
Intelligent Computer and Applications
作者简介
王彬(1996-),男,硕士研究生,主要研究方向:图像处理;张正平(1964-),男,博士,教授,主要研究方向:无线电技术、电信技术、信息处理与控制;陆安江(1978-),男,博士,教授,主要研究方向:优化通信与信息系统。
