动态光照环境下基于轻量网络的磁瓦正反面识别

Recognition of Positive and Negative Sides of Magnetic Tiles Based on Lightweight Network in Dynamic Illumination Environment

在对磁瓦进行分拣封箱时基本依赖传统人工在严格的光照条件下进行分类,需要较大的人力物力的消耗。为提高磁瓦分拣的智能化水平,提出将深度学习引入到磁瓦正反分类当中,利用轻量型网络对磁瓦图像进行正反分类。首先考虑光线角度问题进行磁瓦正反面数据集的采集,并对其进行预处理得到标准图像。然后利用MobileNet对磁瓦正反面进行训练和分类。该系统可以达到99.6%的分类准确率,实验结果表明该方法所用参数量少,对硬件设备需求低,具有实际可行性,对磁瓦分类具有良好的效果。

When sorting and sealing the magnetic tile,it basically depends on the traditional manual classification under strict lighting conditions,which requires a large consumption of manpower and material resources.In order to improve the intelligent level of magnetic tile sorting,it is proposed that depth learning is introduced into the positive and negative classification of magnetic tile,and the lightweight network is used to classify the magnetic tile image.Firstly,the problem of light angle is considered to collect the data set of the positive and negative sides of the magnetic tile,and the standard image is obtained by pre-processing.Then use MobileNet to train and classify the front and back of the magnetic tile.The system can achieve 99.6%classification accuracy.The experimental results show that the method uses a small number of parameters,low demand for hardware equipment,has practical feasibility,and has a good effect on magnetic tile classification.