摘要
基于乙酰胆碱酯酶检测卡与有机磷农药的显色反应,搭建了一套快速获取反应后图像的机器视觉系统,完成了图像的去噪、边缘提取等处理过程,准确提取出检测卡的RGB数值。基于RGB彩色模型,实现了遗传模拟退火算法建立农药残留检测卡R、B颜色特征与农残浓度之间的非线性映射关系。实验证明遗传模拟退火算法对农药残留检测的标定均方根误差小,确定系数高,拟合优势大。
Based on the color reaction of acetylcholinesterase detection card and organophosphorus pesticide,a set of machine vision system for quickly obtaining post-reaction images was constructed.The process of image denoising and edge extraction was completed,and the RGB of the detection card was accurately extracted.Based on the RGB color model,the genetic simulated annealing algorithm was used to establish a nonlinear mapping relationship between the color characteristics of pesticide residue detection cards R and B and the concentration of pesticide residues.The experiment proves that the genetic simulated annealing algorithm has small rms error in calibration of pesticide residue detection,high determination coefficient and large fitting advantage.
作者
丁建军
杨飘
曲令帅
孙超
张明杨
DING Jian-jun;YANG Piao;QU Ling-shuai(Jianghan University, Wuhan, Hubei 430056)
出处
《安徽农业科学》
CAS
2019年第1期253-257,共5页
Journal of Anhui Agricultural Sciences
基金
湖北省知识产权局项目(2017061668)
武汉市市属高校产学研结合项目(CXY201803)
关键词
农药残留
图像处理
遗传模拟退火算法
Pesticide residue
Image processing
Genetic simulated annealing algorithm
