太湖部分地区蓝藻水华频发,对水生态环境和人类健康造成严重威胁.为提升蓝藻水华监测与预警能力,利用已建环太湖蓝藻视频监控资源,基于人工智能图像识别技术,依托 AI 开放平台,开展模型自动化训练和本地化布署;提出表观蓝藻水华强度分析评价体系,对模型识别成果进行加工应用、集成展示,构建环太湖蓝藻图像智能识别系统.结果表明:系统实现监控点表观蓝藻水华分布形态的在线识别与强度分析评价,使蓝藻调查由被动巡查转为主动监测和实时预警,大幅提高太湖蓝藻水华强度监测与预警能力,为其他类似环境监测问题提供可借鉴的解决方案.
Design and application of intelligent algae recognition system for Taihu Lake
The frequent occurrence of cyanobacterial blooms in part of Taihu Lake region poses significant threats to the aquatic ecosystem and public health.To enhance monitoring and early warning capabilities for cyanobacterial blooms,this study utilizes the existing video surveillance resources around Taihu Lake.Leveraging artificial intelligence(AI)image recognition technology and supported by an AI open platform,the system undertakes automated model training and local deployment.Additionally,a cyanobacterial bloom intensity analysis and evaluation framework is proposed to process,apply,and visually integrate the recognition results.The intelligent algae recognition system for Taihu Lake region is thus developed.Results demonstrate that the system enables real-time identification and intensity evaluation of cyanobacterial bloom distribution at monitoring points.This transition from passive inspection to proactive monitoring and early warning significantly enhances the region's monitoring and early warning capabilities regarding cyanobacterial bloom intensity,offering a referable solution for similar environmental monitoring challenges.
万方数据
维普
