[1]杨浩,张通,阳苇丽,等.基于图像处理的雪茄烟叶晾制期间含水率预测模型比较[J].江苏农业学报,2023,(09):1891-1900.[doi:doi:10.3969/j.issn.1000-4440.2023.09.011]
　YANG Hao,ZHANG Tong,YANG Wei-li,et al.Comparison of prediction models for moisture content of cigar tobacco leaves during drying period based on image processing[J].,2023,(09):1891-1900.[doi:doi:10.3969/j.issn.1000-4440.2023.09.011]
点击复制

基于图像处理的雪茄烟叶晾制期间含水率预测模型比较()
分享到：

江苏农业学报[ISSN:1006-6977/CN:61-1281/TN]

卷:
期数:
2023年09期
页码:
1891-1900
栏目:
农业信息工程
出版日期:
2023-12-31

文章信息/Info

Title:
Comparison of prediction models for moisture content of cigar tobacco leaves during drying period based on image processing
作者:
杨浩1; 张通1; 阳苇丽2; 向欢3; 郭仕平4; 刘晓丽1; 张洪淇1; 刘雷1; 刘雅洁1; 杨兴有4; 曾淑华1
(1.四川农业大学农学院,四川成都611130；2.四川省烟草公司达州市公司,四川达州635000;3.四川省烟草公司德阳市公司,四川德阳618400;4.四川省烟草公司,四川成都610017)
Author(s):
YANG Hao1; ZHANG Tong1; YANG Wei-li2; XIANG Huan3; GUO Shi-ping4; LIU Xiao-li1; ZHANG Hong-qi1; LIU Lei1; LIU Ya-jie1; YANG Xing-you4; ZENG Shu-hua1</
(1.Agriculture College of Sichuan Agricultural University, Chengdu 611130, China;2.Sichuan Provincial Tobacco Company Dazhou Branch, Dazhou 635000, China；3. Sichuan Provincial Tobacco Company Deyang Branch, Deyang 618400, China；4. Sichuan Provincial Tobacco Company, Chengdu 610017, China)
关键词:
雪茄烟叶; 含水率预测; 数字图像; BP神经网络; 支持向量机; 极限学习机
Keywords:
cigar tobacco leaves; moisture content prediction; digital images; BP neural network; support vector machine; extreme learning machine
分类号:
TS453
DOI:
doi:10.3969/j.issn.1000-4440.2023.09.011
文献标志码:
A
摘要:
为探索雪茄烟叶晾制期间的水分变化规律并实现含水率的快速准确预测,在雪茄烟叶各个晾制阶段拍摄数字图像,同时以杀青烘干法测定烟叶含水率。提取图像中的颜色和纹理特征作为初始特征,经正交偏最小二乘(OPLS-DA)确定优选特征。分别以初始特征和优选特征为输入,含水率为输出,建立前馈神经网络(BPNN)、遗传算法优化前馈神经网络(GA-BPNN)、支持向量机(SVM)、遗传算法优化支持向量机(GA-SVM)、极限学习机(ELM)、粒子群算法优化极限学习机(PSO-ELM)模型。结果表明,①随着晾制时间推移,烟叶含水率逐渐降低,各阶段含水率差异显著；②优选特征建立的GA-SVM对全晾制阶段的含水率整体预测能力相较于其他模型表现最佳,决定系数(R2)和均方根误差(RMSE)分别为0.969 3、0.044 7；③优选特征建立的GA-SVM对各晾制阶段的含水率预测准确度较高,其中干筋期含水率预测准确度最低,但也高于87.0%。说明,采用OPLS-DA优选的颜色特征、纹理特征建立的GA-SVM可准确预测雪茄烟叶晾制期间的含水率。
Abstract:
To explore the law of moisture change during the drying period of cigar tobacco leaves and realize fast and accurate prediction of moisture content, digital images were taken at each drying stage of cigar tobacco leaves, and the moisture contents of tobacco leaves was determined by method of curing and drying. The color and texture features in the images were extracted as the initial features, and the optimal features were determined by orthogonal partial least squares-discriminant analysis (OPLS-DA). Models of back propagation neural network (BPNN), genetic algorithm optimized BPNN (GA-BPNN), support vector machine (SVM), genetic algorithm optimized SVM (GA-SVM), extreme learning machine (ELM), particle swarm optimized (PSO)-ELM were established with initial features and optimal features as input and moisture content as output, respectively. The results showed that, with the drying time increased, the moisture content of tobacco leaves gradually decreased, and the moisture contents between different stages showed significant differences. Compared with other models, the overall prediction ability of moisture content in the whole drying stage by GA-SVM established by optimized characteristics was the best, coefficient of determination (R2) and root mean squared error (RMSE) were 0.969 3 and 0.044 7， respectively. The prediction accuracy of GA-SVM established by the optimized features for each drying stage was high, and the prediction accuracy of moisture content in dry gluten stage was the lowest, but was also higher than 87.0%. The results indicated that, GA-SVM constructed by the color and texture features optimized by OPLS-DA can predict the moisture content of cigar tobacco leaves during drying stage accurately.

参考文献/References:

[1]杨月先,李宗平,付庆灵,等. 采收和晾制方式对雪茄烟中上部烟叶品质的影响[J]. 烟草科技,2022,55(8):28-34.
[2]国家烟草质量监督检验中心. Y烟草及烟草制品试样的制备和水分的测定-烘箱法:C/T 31-1996 [S]. 北京:中国标准出版社,1996.
[3]中国烟草标准化研究中心. 烟草水分的测定-卡尔费休法:YC/T 165-2003 [S]. 北京:中国标准出版社,2003.
[4]李小玲. 复烤烟叶快速水分测定法和烘箱水分测定法的比较[J]. 工业计量,2006,16(4):28-29.
[5]郭朵朵. 基于机器视觉的烟草烘烤品质检测及控制系统的设计[D]. 杨凌:西北农林科技大学,2017.
[6]HARALICK R M, SHANMUGAM K, DINSTEIN I. Textural features for image classification[J]. IEEE Transactions on Systems Man & Cybernetics,1973,3(6):610-621.
[7]陈飞程,杨懿德,李常军,等. 基于图像信息的烤烟烘烤过程中烟叶含水率预测[J]. 西南农业学报,2021,34(11):2378-2384.
[8]杜海娜,孟令峰,王松峰,等. 基于机器学习的密集烘烤过程烟叶失水率预测模型对比[J]. 烟草科技,2022,55(9):81-88.
[9]李增盛,孟令峰,王松峰,等. 基于图像处理的烟叶烘烤阶段判别模型优选[J]. 中国烟草学报,2022,28(2):65-76.
[10]GHOSH A, ROY P. An automated model for leaf image-based plant recognition: an optimal feature-based machine learning approach[J]. Innovations in Systems and Software Engineering,2022,3:1-14.
[11]宋佳. 基于图像处理的生菜叶片含水率检测研究及Android平台的实现[D]. 镇江:江苏大学,2017.
[12]张志良. 植物生理学实验指导[M]. 北京:高等教育出版社,1990:1-10.
[13] MEYER G E, HINDMAN T W, LAKSMI K. Machine vision detection parameters for plant species identification[J]. Proceedings of Spie the International Society for Optical Engineering,1999,1.DOI:10.1117/12.336896.
[14] WOEBBECKE D M, MEYER G E, VON BARGEN K, et al. Color indices for weed identification under various soil, residue, and lighting conditions[J]. Transactions of the ASAE,1995,38(1):259-269.
[15] GUIJARRO M, PAJARES G, RIOMOROS I, et al. Automatic segmentation of relevant textures in agricultural images[J]. Computers and Electronics in Agriculture,2010,75(1):75-83.
[16] LOUHAICHI M, BORMAN M M, JOHNSON D E. Spatially located platform and aerial photography for documentation of grazing impacts on wheat[J]. Geocarto International,2001,16(1). DOI:10.1080/10106040108542184.
[17] LIANGSHENG Y Q Y H. Estimation of leaf area index of sugarcane using crop surface model based on UAV image[J]. Journal of Agricultural Engineering,2017,33(8):104-111.
[18] WANG Y, WANG D, ZHANG G, et al. Estimating nitrogen status of rice using the image segmentation of G-R thresholding method[J]. Field Crops Research,2013,149:33-39.
[19] SMITS B. An RGB-to-spectrum conversion for reflectances[J]. Journal of Graphics Tools,2000,1:1-10.
[20]白竣文,田潇瑜,马海乐. 基于BP神经网络的葡萄气体射流冲击干燥含水率预测[J]. 现代食品科技,2016,32(12):198-203.
[21] VAN HENTEN E J. Non-destructive crop measurement by image processing for crop growth control [J]. Journal of Agricultural Engineering Reassearch,1995,61:97-105.
[22]程晓婷. 基于番茄叶片图像的土壤含水率研究[D]. 银川:宁夏大学,2021.
[23]徐腾飞,韩文霆,孙瑜. 基于图像处理的玉米叶片含水率诊断方法研究[J]. 干旱地区农业研究,2013,31(1):95-100.
[24]王甜,王雪峰,刘嘉政. 基于RFE＿RF算法的幼龄沉香叶片含水率预估模型[J]. 南京林业大学学报(自然科学版),2022,46(4):177-184.
[25]段史江,宋朝鹏,马力,等. 基于图像处理的烘烤过程中烟叶含水量检测[J]. 西北农林科技大学学报(自然科学版),2012,40(5):74-80.
[26]詹攀,谢守勇,刘军,等. 基于支持向量机回归的鲜烟叶含水量预测模型[J]. 西南大学学报(自然科学版),2016,38(4):165-170.
[27]张娟利,韩文霆,杨六生,等. 基于图像处理烤烟鲜烟叶含水率诊断研究[J]. 现代农业科技,2017(9):3-5.
[28]赵晨,王爱霞,范宁波,等. 基于CIE-Lab的雪茄烟叶晾制过程水分与颜色相关性研究[J]. 湖南文理学院学报(自然科学版),2022,34(2):60-66.
[29]范宁波. 雪茄烟叶晾制过程中颜色及相关生理变化特性分析[D]. 郑州:河南农业大学,2021.
[30]谢滨瑶,祝诗平,黄华. 基于BPNN和SVM的烟叶成熟度鉴别模型[J]. 中国烟草学报,2019,25(1):45-50.

备注/Memo

备注/Memo:
收稿日期:2022-11-24基金项目:四川省烟草公司科技项目(SCYC202121)作者简介:杨浩(1997-),云南昭通人,硕士研究生,研究方向为烟草栽培。(E-mail)2551958014@qq.com通讯作者:刘雅洁,(E-mail)31917272@qq.com；杨兴有,(E-mail)tobaccoboy@163.com；曾淑华,(E-mail)zshgsp@163.com
更新日期/Last Update: 2024-01-15