摘要 介绍了RFID技术和图像处理技术,研究了基于图像处理和卷积神经网络的水果识别与分类模型,设计了一套水果目标物识别处理系统,可以实现对苹果、梨子、水蜜桃和香蕉等4种水果的识别与分类。实验结果表明:卷积神经网络模型采用自主学习型的网络,识别错误率只有0.93%,相比其他传统模型分类错误率降低较多,证明了卷积神经网络模型性能优良,具有较好的可行性。 Firstly,it introduces the technology of RFID and image processing,studies the fruit recognition and classification model based on image processing and convolution neural network,realizes a set of fruit target recognition and processing system,which can recognize and classify four kinds of fruits,such as apple,pear,peach and banana.The experimental results show that the recognition error rate of the convolutional neural network model is only 0.93 percentage points,which is much lower than that of other traditional models.It proves that the convolutional neural network model has good performance and high validity and feasibility.
机构地区 河南农业职业学院
出处 《农机化研究》 北大核心 2021年第4期42-46,共5页 Journal of Agricultural Mechanization Research
基金 河南省高等职业院校创新发展行动计划项目(XM0156)。
关键词 水果 识别 分类 RFID技术 图像处理 卷积神经网络 fruit recognition classification RFID technology image processing convolution neural network
