基于机器视觉和工艺参数的针芽形绿茶外形品质评价
发布时间：2019-01-04 15:47
【摘要】：外形是针芽形绿茶的关键感官评价指标,通常依据色泽、条形、嫩度和匀整度等表象特征进行人工评审,难以做到精准、客观和量化评价。本文以自动化生产线机制的针芽形绿茶为研究对象,基于茶叶品质、形成工艺和视觉形态等内外因素,构建了外形品质的智能感官评价方法。首先,在线采集在制品的17个机制工艺参数和成品茶的图像,进行图像特征提取,选取9个颜色特征和6个纹理特征。进而,通过与专家感官评分进行关联分析,明确了与感官品质显著相关的特征变量。为获取高效的评价模型,采用偏最小二乘法(PLS)、极限学习机(ELM)和强预测器集成算法(ELM-Ada Boost)3种多元校正方法,分别建立了基于工艺或图像特征的针芽形绿茶外形感官的量化评价模型。建模结果表明,基于图像特征建立的ELM-Ada Boost模型(Rp=0.892,RPD大于2),其预测性能优于其他模型,且具有更小的RMSEP(0.874)、Bias(-0.148)、SEP(0.226)和CV(0.018)值。同时,非线性模型的预测性能均高于PLS线性模型,能更好地表征工艺参数、图像信息与感官评分之间的解析关系,且建模速度更快(0.014~0.281 s)。而Ada Boost法作为一种混合迭代算法,能进一步提升ELM模型的精度和泛化能力。结果表明,基于机器视觉和工艺评价针芽形绿茶外形品质是可行的,为拓展茶叶感官品质评价方法和专家工艺决策支持系统研制,提供理论依据和数据支撑。
[Abstract]:Shape is the key sensory evaluation index of needle bud green tea. It is difficult to evaluate accurately objectively and quantitatively according to the color strip tenderness and evenness of green tea. Based on the internal and external factors such as tea quality, forming process and visual form, the intelligent sensory evaluation method for the appearance quality of needle bud green tea with automatic production line mechanism was established in this paper. Firstly, 17 process parameters of in-process products and images of tea products were collected online, and 9 color features and 6 texture features were selected for image feature extraction. Furthermore, through the correlation analysis with the expert sensory score, the characteristic variables which are significantly related to the sensory quality are identified. In order to obtain an efficient evaluation model, three multivariate correction methods, partial least square (PLS),) extreme learning machine (ELM) and strong predictor integrated algorithm (ELM-Ada Boost), are adopted. A quantitative sensory evaluation model of needle bud green tea was established based on process or image features. The modeling results show that the prediction performance of the ELM-Ada Boost model (Rp=0.892,RPD > 2) based on image features is better than that of other models, and it has a smaller RMSEP (0.874), Bias (- 0.148). SEP (0.226) and CV (0.018) values. At the same time, the prediction performance of the nonlinear model is better than that of the PLS linear model, which can better represent the analytical relationship between the process parameters, the image information and the sensory score, and the modeling speed is faster (0.014 / 0.281 s). As a hybrid iterative algorithm, Ada Boost method can further improve the accuracy and generalization ability of ELM model. The results showed that it was feasible to evaluate the shape quality of needle bud green tea based on machine vision and technology, which provided theoretical basis and data support for the development of tea sensory quality evaluation method and expert process decision support system.
【作者单位】： 江苏大学食品科学与食品工程学院;中国农业科学院茶叶研究所;哥本哈根大学食品科学系;武义县农业局;
【基金】：国家自然科学基金项目(31271875) 浙江省自然科学基金项目(Y16C160009) 中央级公益性科研院所基本科研业务费专项(1610212016018)
【分类号】：TP391.41;TS272.51
本文编号：2400486
[Abstract]:Shape is the key sensory evaluation index of needle bud green tea. It is difficult to evaluate accurately objectively and quantitatively according to the color strip tenderness and evenness of green tea. Based on the internal and external factors such as tea quality, forming process and visual form, the intelligent sensory evaluation method for the appearance quality of needle bud green tea with automatic production line mechanism was established in this paper. Firstly, 17 process parameters of in-process products and images of tea products were collected online, and 9 color features and 6 texture features were selected for image feature extraction. Furthermore, through the correlation analysis with the expert sensory score, the characteristic variables which are significantly related to the sensory quality are identified. In order to obtain an efficient evaluation model, three multivariate correction methods, partial least square (PLS),) extreme learning machine (ELM) and strong predictor integrated algorithm (ELM-Ada Boost), are adopted. A quantitative sensory evaluation model of needle bud green tea was established based on process or image features. The modeling results show that the prediction performance of the ELM-Ada Boost model (Rp=0.892,RPD > 2) based on image features is better than that of other models, and it has a smaller RMSEP (0.874), Bias (- 0.148). SEP (0.226) and CV (0.018) values. At the same time, the prediction performance of the nonlinear model is better than that of the PLS linear model, which can better represent the analytical relationship between the process parameters, the image information and the sensory score, and the modeling speed is faster (0.014 / 0.281 s). As a hybrid iterative algorithm, Ada Boost method can further improve the accuracy and generalization ability of ELM model. The results showed that it was feasible to evaluate the shape quality of needle bud green tea based on machine vision and technology, which provided theoretical basis and data support for the development of tea sensory quality evaluation method and expert process decision support system.
【作者单位】： 江苏大学食品科学与食品工程学院;中国农业科学院茶叶研究所;哥本哈根大学食品科学系;武义县农业局;
【基金】：国家自然科学基金项目(31271875) 浙江省自然科学基金项目(Y16C160009) 中央级公益性科研院所基本科研业务费专项(1610212016018)
【分类号】：TP391.41;TS272.51
【相似文献】
相关期刊论文 前1条
1 朱宏凯;董春旺;张坚强;刘飞;叶阳;;针芽形名优绿茶滚筒杀青工艺优化研究[J];中国农学通报;2014年18期
相关会议论文 前1条
1 董春旺;张坚强;洪文娟;何华锋;朱宏凯;刘飞;叶阳;;针芽形绿茶外形品质智能化感官评价方法研究[A];第十六届中国科协年会——分12茶学青年科学家论坛论文集[C];2014年
相关硕士学位论文 前1条
1 陈根生;针芽形名优绿茶连续化加工关键技术研究[D];浙江大学;2016年
，本文编号：2400486
本文链接：https://www.wllwen.com/kejilunwen/ruanjiangongchenglunwen/2400486.html
