摘要
本文是基于深度学习对环境地表水源质量等级进行分类研究及应用,通过进一步优化深度学习算法,选择合适的深度学习CNN网络模型,包括但不限于VGG16、DenseNet、Resnet50、AlexNet、Faster R-CNN等中的图像识别技术,分别对不同的模型选择合适的参数,进行训练和调参,找到最合适的训练模型和参数设置。联合环境水质检测技术,将地表水源水质检测数据和地表水现状图片相结合,进行关联性研究分析,建立水质参数检测结果与水质水平之间复杂的、非线性的因果关系,从而建立地表水水源质量等级分类系统,实现通过水质现状图片快速、准确地对未知地表水进行水质的分类识别。
This paper classifies and applies the quality grade of environmental surface water based on deep learning.By further optimizing the deep learning algorithm,appropriate deep learning CNN network model is selected.Including but not limited to image recognition technologies in VGG16,DenseNet,Resnet50,AlexNet,Faster R-CNN,etc.,select appropriate parameters for different models,conduct training and parameter tuning,and find the most appropriate training model and parameter setting.Combined with environmental water quality detection technology,the surface water quality detection data and surface water status picture are combined to conduct correlation research and analysis,and the complex and non-linear causal relationship between the detection results of water quality parameters and water quality level is established,so as to establish the surface water quality classification system.It can classify and identify unknown surface water quality quickly and accurately by picture of water quality status.
作者
周美姣
董俐香
吴佳
顾桔
Zhou Meijiao;Dong Lixiang;Wu Jia;Gu Ju(Qingshan Lvshui(Jiangsu)Inspection and Testing Co.,Ltd.,Changzhou 213000,China)
出处
《皮革制作与环保科技》
2022年第9期190-191,194,共3页
Leather Manufacture and Environmental Technology
关键词
深度学习
地表水
质量等级
分类技术
deep learning
surface water
quality grade
classification techniques
