基于卷积神经网络的物品识别系统设计与实现

摘要

In order to obtain the information of the unknown goods quickly,the separation of server and client architecture is adopt to design and implement a object recognition system on the computer side and Android mobile terminal based on convolution neural network.Firstly,it pre-processed the local pictures or cell phone photos uploaded from the client side,and then built a convolutional neural network by using the pre training model to extract image features and classified them.Finally,it returned the result to the client via a network connection to complete the object recognition.The experimental results show that the system has good online recognition ability,and this will greatly improves the way people access to information,it is more convenient and efficient.%为快速准确地获取未知物品的信息,采用服务端和客户端相分离的架构,设计并实现基于卷积神经网络的电脑端和Android手机端物品识别系统.首先对客户端上传的本地图片或者手机拍摄照片进行预处理,然后利用预训练模型搭建卷积神经网络,提取图像特征并分类,最后将识别结果通过网络连接返回至客户端,完成物品的识别.实验结果表明,该系统拥有良好的在线识别能力,这将极大地改进人们获取信息的方式,更加方便和高效.