一种酶标板培养孔颜色特征提取的图像检测方法  

Image Detection Method for Extracting Color Features of ELISA Plate Wells

作　　者：邹盛 赵立宏[1] ZOU Sheng;ZHAO Lihong(School of Mechanical Engineering,University of South China,Hengyang 421001,China)

机构地区：[1]南华大学机械工程学院,湖南衡阳421001

出　　处：《自动化仪表》2018年第12期9-12,17,共5页Process Automation Instrumentation

摘　　要：针对现阶段酶标(ELISA)板培养孔菌体生长情况的筛选判断检测方法存在效率低、自动化程度低等问题,提出了采用机器视觉算法对酶标板培养孔菌液颜色识别的菌体生长检测方法。对酶标板孔菌体在不同生长时期、菌液浑浊度不同产生的颜色差异进行调查研究,在OpenCV平台对摄像头采集的酶标板图片信息进行预处理和边缘检测,寻找酶标板的最小包围矩形确定目标区域,选取培养孔孔位区域进行颜色计算获取判断依据。考虑到摄像头在垂直方向拍摄时酶标板四周的孔位区域会发生变形,提出了在不同孔位区域选取合适区域计算颜色特征的试验方案,可有效减小培养孔颜色特征提取的误差,为实现在微生物培养中完成菌体的自动化筛选奠定了基础。Aiming at the problems of low efficiency and low degree of automation in the screening and judging method of the growth of the microbial cells in the current stage of the enzyme linked immunosorbent assay(ELISA)plate,a method for detecting the growth of the cells by using the machine vision algorithm for the color identification of the cultured cells is proposed.The color difference caused by the difference of the turbidity of the bacteria in the different stages of the growth of the cells is investigated,the image information of the microplate collected from the camera is used for pre-processing and edge detection on the OpenCV platform.The minimum enveloping rectangle of the microplate is used to determine the target area,and the area of the culture hole is selected for color calculation to obtain the judgment basis.Considering that the hole area around the microplate is deformed when the camera is photographed in the vertical direction,an experimental scheme for selecting the appropriate area to calculate the color features in different hole areas is proposed,which can effectively reduce the error of the color feature extraction of the culture hole,to provide the basis of automated screening of the cells in microbial culture.

关 键 词：微生物培养 机器视觉 酶标板 培养定位矫正 颜色提取 OPENCV 

分 类 号：TH79[机械工程—仪器科学与技术] TP391[机械工程—精密仪器及机械]