摘要
提出了一种利用图像处理技术快速检测圆褐固氮菌浓度的方法。通过对圆褐固氮菌微视图像的采集、预处理、分割和特征提取,运用支持向量机(SVM)进行识别、分类和计数,获取圆褐固氮菌的浓度,实现了圆褐固氮菌活性的快速检测。该方法检测精度较高,与精确的人工计数相比,其相对误差小于4%,但比人工计数省时省力;与专用检测设备相比,其检测成本低。
A measuring method using image processing techniques for the rapid detection of Azotobacter chroococcum was put forward.Based on the microscopic image collection,preprocessing,segmentation and feature extraction,as well as the support vector machine（SVM） identification,classification and counting,Azotobacter chroococcum concentration density was acquired.The rapid Azotobacter chroococcum activity detection was realized.Experiments showed that detection precision of the proposed method was higher,and its relative error was less than 4% compared with precisely manual counting.Meanwhile,the detection method was more time saving.As for the other special detecting devices,the measuring detection cost was much lower.
出处
《农业机械学报》
EI
CAS
CSCD
北大核心
2012年第10期174-178,共5页
Transactions of the Chinese Society for Agricultural Machinery
基金
江苏省高校优势学科建设工程资助项目(苏政办发(2011)6号)
关键词
圆褐固氮菌
图像处理
支持向量机
检测
Azotobacter chroococcum
Image processing
Support vector machine
Detection
