首页|基于深度学习的电梯按键的识别方法

基于深度学习的电梯按键的识别方法

扫码查看
针对目前因数据集的缺陷导致对电梯按键识别准确率不高的问题,提出了一种用于识别电梯按键的方法.该方法将多个角度收集的电梯按键图像作为数据集;通过形态学运算处理降低了数据噪声,凸显了数据特征;建立了一种训练模型来完成对图像的训练.该方法在训练和实验阶段识别准确率分别为95.90%和97.17%.
Recognition method of elevator buttons based on deep learning
The accuracy of elevator key recognition is not high due to the defect of data set,so a method for ele-vator key recognition is proposed.The method uses multi-angle collection of elevator key images as a data set.Through morphological processing,data noise is reduced and data characteristics are highlighted.A training structure is established to complete the training of pictures.The recognition accuracy of this method is 95.90%and 97.17%respectively in the training and experimental stages,and the recognition of elevator keys can a-chieve good results.

deep learningneural networkdataset establishmentelevator button recognitionimage processing

李伊童、曹旭阳、蒲长林

展开 >

大连理工大学机械工程学院,辽宁 大连 116024

深度学习 神经网络 数据集建立 电梯按键识别 图像处理

2025

机械设计与制造工程
南京东南大学出版社有限公司

机械设计与制造工程

影响因子:0.387
ISSN:1672-1616
年,卷(期):2025.54(1)