李德林¹，张宿义^1*，张 程¹，杨丽明²，王 海¹，赵 杰¹，侯 望¹，顾浩田¹，李璧君¹，曾智娟¹

（1.泸州老窖股份有限公司，四川泸州 646000；

2.浙江中控技术股份有限公司，浙江杭州 310000）

摘要：浓香型白酒生产过程中发酵糟醅质量的好坏对于基酒品质和产量都有重大影响。传统发酵糟醅质量的鉴别主要依靠车间技术人员眼看、手捏等方式，受经验及主观判断差异影响较大，无法保障生产配料和基酒品质的稳定性。本文提出了一种基于计算机图像识别技术的快速发酵糟醅形态自动识别方法。

关键词：浓香型白酒；糟醅；图像识别技术

中图分类号：TS262.31   文献标识码：A   文章编号：1674-506X（2021）05-0099-0005

Research on Morphology Recognition Method of Luzhou-flavor Baijiu Fermented Grains Based on Image Recognition Technology

LI De-lin¹, ZHANG Su-yi^1*, ZHANG Cheng¹, YANG Li-ming², WANG Hai¹, ZHAO Jie¹, HOU Wang¹, GU Hao-tian¹, LI Bi-jun¹, ZENG Zhi-juan¹

（1.Luzhou Laojiao Co., Ltd., Luzhou Sichuan 646000, China;

2.Zhejiang Supcon Technology Co., Ltd., Hangzhou Zhejiang 310000, China）

Abstract：The quality of fermented grains during Luzhou flavor Baijiu liquor production has great influence on microbial community, quality and yield of base liquor. Quality of fermented grains mainly depends on the comprehensive judgment of sensory (watching and touching) in traditional identification methods, which is greatly affected by personal experience and subjective judgment. The stability of production ingredients and base wine quality cannot be guaranteed. This paper proposes a rapid automatic identification method of fermented grains shape based on computer image recognition technology.

Keywords：Luzhou flavor Baijiu; fermented grains; image recognition technology

doi：10.3969/j.issn.1674-506X.2021.06-016

收稿日期：2021-06-07

基金项目：2018年智能制造综合标准化与新模式应用项目《固态法白酒生产数字化车间新模式应用》；固态酿造泸州市科技创新团队项目。

作者简介：李德林（1986-），男，硕士，工程师。研究方向：酿酒技术。

*通信作者

引文格式：李德林,张宿义,张程,等.基于图像的浓香型白酒发酵糟醅形态识别方法研究[J].食品与发酵科技,2021,57(6):99-103.