摘要
随着模式识别领域不断的发展,图像识别作为该领域中的典型应用,受到了众多学者的关注,尤其是在图像快速识别方面。对此,该文将从深度学习的概述入手,选择卷积神经网络算法应用于水果图像识别中。通过对卷积神经网络算法的网络结构以及卷积神经网络训练过程的详细分析,采用卷积神经网络架构设计、分类器设计建构了基于卷积神经网络的水果图像识别系统。并将识别结果与传统水果图像识别结果进行对比验证,为水果图像识别领域提供参考。
With continuous development of pattern recognition,image recognition,as a typical application in this field,had attracted attention of many scholars,especially in the area of fast image recognition. Convolution neural network algorithm was applied to fruit image recognition based on deep learning. By analyzing network structure of convolution neural network algorithm and training process of convolution neural network,a fruit image recognition system based on convolution neural network was constructed by using convolution neural network architecture design and classifier design. Recognition results were compared with traditional fruit image recognition results,which provided a reference for field of fruit image recognition.
作者
罗琪
LUO Qi(Weinan Normal University, Weinan Shaanxi 714099, China)
出处
《农业工程》
2018年第10期31-34,共4页
AGRICULTURAL ENGINEERING
关键词
深度学习
水果图像识别
卷积神经网络
deep learning
fruit image recognition
convolution neural network