净水技术2026，Vol.45Issue(3)：188-196,204,10.DOI:10.15890/j.cnki.jsjs.2026.03.020
基于计算机视觉与深度神经网络的数智化混凝试验
Digital-Intelligent Coagulation Experiment Based on Computed Vision and Deep Neural Network
摘要
Abstract
[Objective]With the rapid development of emerging technologies and the deepening integration of interdisciplinary fields,water treatment education faces new opportunities and challenges.This paper aims to enhance students' understanding of the flocculation process by introducing advanced technological means and to explore the potential application of computer vision and deep learning in water treatment education.[Methods]The computer vision technology involved in this paper applied computer algorithms to analyze the pixel matrix of images and extract features.During the coagulation experiment,images of flocs at specific coagulation stages were captured using photography.Image analysis software was then used to quantitatively calculate characteristic parameters of the flocs,such as projected area and perimeter,and to summarize their correlation with coagulation effectiveness.Another approach for floc characteristic analysis involved employing deep neural network model algorithms to achieve intelligent recognition of floc images,systematically investigating their association patterns with coagulation experimental results.[Results]Characteristic parameters of flocs,such as size and fractal dimension obtained through computer vision analysis showed clear correlations with treatment effectiveness,providing a vivid interpretation of coagulation mechanisms.Furthermore,deep neural network algorithms achieved accurate recognition of floc characteristics.The excellent analytical capability of artificial intelligence algorithms for floc images stimulated students' learning enthusiasm.This teaching method enabled students' understanding of floc morphological characteristics and flocculation kinetics to progress from qualitative observation to rational cognition,significantly deepening their comprehension of flocculation dynamics and coagulation mechanisms.[Conclusion]This paper provides a demonstration case for the application of computer vision and deep neural network technologies in water treatment education,laying a foundation for cultivating interdisciplinary innovative talents.关键词
实践教学/混凝试验/絮凝形态学/计算机视觉/深度学习Key words
practice teaching/coagulation experiment/flocculation morphology/computed vision/deep learning分类
资源环境引用本文复制引用
盛力,史俊,唐贤春,陆志波..基于计算机视觉与深度神经网络的数智化混凝试验[J].净水技术,2026,45(3):188-196,204,10.基金项目
同济大学实验教学改革专项基金项目(0400104194) （0400104194）
