摘要
机器视觉检测技术的普及能够有效提高企业陶瓷的生产效率,然而目前机器视觉技术在日用陶瓷检测领域的应用较少。其主要原因是:基于陶瓷的机器视觉在线检测存在一个典型问题——陶瓷眩光。此现象会严重影响陶瓷图像的采集和后续的智能处理。本论文针对陶瓷表面存在眩光的问题,提出一种基于同态滤波的陶瓷防眩光处理技术。本算法不仅可以解决眩光造成的细节不清晰问题,同时还能较清晰还原陶瓷图像的细节。实验结果表明,本文提出的防眩光算法具有较好的实用性。
With the development of machine vision detection techniques, the efficiency of product production and inspection is greatly improved. However, machine vision techniques are not widely employed in ceramic field. One of the critical reasons is that ceramic glare may significantly affect the inspection accuracy. To prevent glare from ceramic surface, this paper proposes a novel homomorphic filtering-based anti-glare scheme. As a pre-processing phase, this algorithm can effectively remove ceramic glare and facilitate high-quality ceramic image acquirement. The experimental results show that the proposed anti-glare algorithm has desired practicability.
作者
王俊祥
张影
胡欢
黄霖
朱永红
WANG Junxiang;ZHANG Ying;HU Huan;HUANG Lin;ZHU Yonghong(Jingdezhen Ceramic Institute, Jingdezhen 333403, Jiangxi, China;Government and Enterprise Customers Department,Jingdezhen Telecom Company, Jingdezhen 333000, Jiangxi, China)
出处
《陶瓷学报》
CAS
北大核心
2019年第1期62-66,共5页
Journal of Ceramics
基金
国家自然科学基金地区基金项目(61762054
61563022)
江西省自然科学基金重大项目(20161ACB21009
20152ACB20009)
江西省杰出青年人才资助计划(20171BCB23072)
江西省教育厅科技项目(GJJ1707619)
国家公派留学基金项目(201708360076)