摘要
水中的氨氮含量是衡量饮用水水质的重要指标,水的浊度也是判断水质最直观的一项指标。但在现有技术成果条件下,检测水的氨氮含量及浊度的相关方法的便利性及快捷性尚存不足。随着人工智能的快速发展,通过拍摄水源并得出水质检测结果是一种便捷可行的方法。该方法基于颜色特征值分析和决策树模型机器学习,可分类并处理大量的水样集。结果表明,通过构建水样图像RGB值与水体浑浊度及水体氨氮浓度之间的相关性模型,可同时检测水样图像中水体的浑浊度和氨氮浓度,检测精度可达到90.24%。
The ammonia content of water is an important indicator of drinking water quality,while the turbidity of water is also one of the most intuitive indicators of water quality.However,under the existing technical achievements,the convenience and speed of the methods for detecting the ammonia content and turbidity of water are still lacking.With the rapid development of artificial intelligence,it is a convenient and feasible method to photograph water sources and produce water quality test results.The method is based on colour eigenvalue analysis and decision tree model machine learning,and can classify and process large sets of water samples.The results show that by building a correlation model between the RGB values of the water sample images and the turbidity and ammonia concentration of the water body,the turbidity and ammonia concentration of the water body in the water sample images can be detected simultaneously with an accuracy of 90.24%.
作者
彭子康
张芹
陈世航
王辉华
吴广飞
PENG Zikang;ZHANG Qin;CHEN Shihang;WANG Huihua;WU Guangfei(School of International Education,Nanchang Hangkong University,Nanchang,Jiangxi 330063,China)
出处
《自动化应用》
2023年第10期188-191,共4页
Automation Application
基金
南昌航空大学测试与光电工程学院院级课题(2022GJCG048)。
关键词
水质检测
机器学习
图像颜色特征
浑浊度
氨氮浓度
water quality detection
machine learning
image color feature
turbidity
ammonia nitrogen concentration
