摘要
Abstract
In water environment detection and water quality assessment,phytoplankton is a key indicator organism,and its de-tection results are of great significance for analyzing ecological changes and early warning of water eutrophication.In this pa-per,two methods of intelligent detection and manual detection of phytoplankton were used to quantitatively evaluate the practi-cability and operability of the two detection methods by comparing and analyzing the characteristics of phytoplankton communi-ty in Xianghongdian reservoir.A total of 53 species of algae in 6 phyla were identified by the two methods,and 46 species in 5 phyla and 35 species in 5 phyla were detected by intelligent detection and manual detection.Artificial detection of species i-dentification was more stable,the estimation of phytoplankton density by intelligent detection was generally high,and there were errors in the identification of dominant species.The comprehensive evaluation grade of phytoplankton was Ⅲ in front of the dam,IV in the center of the reservoir,andⅢ in front of the dam and in the center of the reservoir.Intelligent detection has insufficient accuracy in species(genus)level identification,population algae counting and special species identification.Man-ual detection is irreplaceable in complex sample identification.In the future,it is necessary to promote the deep integration of the two.关键词
浮游植物/浮游植物密度/显微镜/深度神经网络技术/图像识别/响洪甸水库Key words
phytoplankton/phytoplankton density/microscope/deep neural network technology/image recognition/Xianghong-dian reservoir分类
资源环境
