摘要
针对模具制品射出成型过程中因制品滞留在模腔中造成的夹模问题,提出集数字图像采集与处理以及模式识别等技术的解决方案,实现在注塑行业成型模具的监视系统。运用自适应阈值分割算法进行图像处理,运用图像差影检测算法进行模式匹配比较分析,并利用基于搜索空间与相似度测量技术的位置补偿算法,通过自动位置补偿将采集到的图像进行矫正。结果表明,生产制品的品质合格品率由90%提高到99%,注塑成型机的使用率由75%提高到96%。
Aiming at the problem of clamping mold because of mold products staying in mold cavity in the process of injection molding,this paper proposes a solution,which applies digital image acquisition,processing and pattern recognition,and the mold monitoring system is developed for plastic injection to solve.Images are processed by applying self-adaptive threshold segmentation algorithm,pattern matching is analyzed by applying.Image difference detection algorithm,the compensation algorithm with search space and similarity measurement technique is used to correct the tilted image automatically by using the edge detection technique.Result shows that the ratio of qualified products is improved from 90% to 99%,the utilization rate of injection molding machines is increased from 75% to 96%.
出处
《计算机工程》
CAS
CSCD
北大核心
2010年第9期238-239,242,共3页
Computer Engineering
基金
福建省2006年科技计划基金资助项目(2006H0035)
福建省2006年高等学校新世纪优秀人才支持计划基金资助项目(NCET)
厦门市2008年科技计划基金资助项目(3502Z20081098)
关键词
位置补偿
图像分割
边缘检测
图像识别
模具监视
position compensation
image segmentation
edge detection
image recognition
mold monitoring
