Dongetal.JZhejiangUniv-Sci.PDFVIP

Dong et al. / J Zhejiang Univ-Sci B (Biomed Biotechnol) in press 1 Journal of Zhejiang University-SCIENCE B (Biomedicine Biotechnology) ISSN 1673-1581 (Print); ISSN 1862-1783 (Online) /jzus; E-mail: jzus@ Sensory quality evaluation for appearance of needle-shaped green * tea based on computer vision and nonlinear tools 1,2 2 1 Chun-wang DONG , Hong-kai ZHU , Jie-wen ZHAO , Yong-wen JIANG1,2, Hai-bo YUAN†‡2, Quansheng CHEN†‡1 1 ( School of Food and Biological Engineering, Jiangsu University, Zhenjiang 212013, China) (2 Tea Research Institute, the Chinese Academy of Agricultural Sciences, Hangzhou 310008, China) †E-mail: 8517809@; q.s.chen@ d Received Oct. 1, 2016; Revision accepted Dec. 16, 2016; Crosschecked Jan. 6, 2017 Abstract: As the key evaluation index of needle-shaped green tea quality, the appearance of tea is hard to be eval- uated accurately, objectively and quantitatively by human sense organs. Based on computer vision and image pro- e cessing technologies, this paper extracted the color and texture characteristics of the tea sample images, analyzed the correlation with the sensory evaluation by experts, and selected 10 remarkably correlated characteristic variables. After that, it adopted a hybrid algorithm of Adaboost algorithm and ELM neural network, and successfully established a t nonlinear pr