当前位置>主页 > 期刊在线 > 智能制造 >

智能制造22年14期

基于自适应策略的模具监视器算法研究
易非凡,石滨萌
(深圳职业技术学院 粤港澳大湾区人工智能应用技术研究院,广东 深圳 518055)

摘  要:市面上常见的模具监视器算法往往需要专业技术人员根据现场情况,设置大量的参数(如检测阈值、判定阈值、预处理参数等),不利于普通工人作业。针对这一问题,提出了一种无须设置参数,可自适应图像内容的模具监视器算法。首先,采集空模穴的模板图像及满模穴的对比图像;其次,基于 ROI 的直方图分布,采用自适应策略获得算法的相关参数和差异标准阈值;最后,基于获取的参数和阈值,对源图像的 ROI 区域进行检测判断。在实际生产的注塑机上进行对比实验,结果表明,算法能够在无人为设置参数的情况下,达到漏检率低、误检率低的检测标准。


关键词:无参数;自适应;检测算法;模具监视器



DOI:10.19850/j.cnki.2096-4706.2022.014.033


中图分类号:TP391.4                                      文献标识码:A                                 文章编号:2096-4706(2022)14-0142-04


Research on Mold Monitor Algorithm Based on Adaptive Strategy

YI Feifan, SHI Binmeng

(Research Institute of Applied Artificial Intelligence Technology of the Guangdong-Hong Kong-Macao Greater Bay Area, Shenzhen Polytechnic, Shenzhen 518055, China)

Abstract: Common mold monitor algorithms on the market often require professional technicians to set a large number of parameters (such as detection thresholds, judgment thresholds, preprocessing parameters and so on) according to on-site conditions, which is unfavorable for ordinary workers to work. Aiming at this problem, a mold monitor algorithm that can adapt to the image content without setting parameters is proposed. Firstly, it collects the template image of the empty cavity and the comparison image of the full cavity. Secondly, based on the histogram distribution of ROI, it uses the adaptive strategy to obtain the relevant parameters and the difference standard threshold of the algorithm. Finally, based on the obtained parameters and thresholds, it detects and judges the ROI region of the source image. The comparison experiment is carried out on the injection molding machine of actual production, and the results show that the algorithm can achieve the detection standard of low missed detection rate and low false detection rate on the situation without setting parameters manually.

Keywords: no parameter; adaptive; detection algorithm; mold monitor


参考文献:

[1] SIMONYAN K,ZISSERMAN A. Very Deep Convolutional Networks for Large-Scale Image Recognition [J/OL].arXiv:1409.1556 [cs.CV].[2022-04-18]. https://arxiv.org/abs/1409.1556. 

[2] REDMON J,DIVVALA S,GIRSHICK R,et al. YouOnly look Once: Unified, real-Time object Detection [C]//2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).Las Vegas:IEEE,2016:779-788.

[3] 姜旭飞,邢科新,董辉,等 . 模具监视系统中的图像快速配准算法 [J]. 化工自动化及仪表,2012,39(11):1468-1472+1535.

[4] 姚俊峰,詹长明,HUNG C C. 基于位置补偿技术的模具监视系统 [J]. 计算机工程,2010,36(9):238-239+242.

[5] 李运秀 . 一种实现模具监视器的图像处理软件算法:CN201511010200.6 [P].2017-07-07.

[6] 游旭新 . 模具监视系统及方法:CN201910538875.X[P].2019-09-20.

[7] 王海丰,李娟 . 一种基于直方图的亮度评估探究 [J]. 信息与电脑(理论版),2021,33(22):124-126.

[8] 丁国绅,乔延利,易维宁,等 . 基于光谱图像空间的改进SIFT 特征提取与匹配 [J]. 北京理工大学学报,2022,42(2):192-199.


作者简介:易非凡(1994.03—)男,汉族,湖南邵东人,科研助理,硕士,研究方向:机器学习、图像处理;通讯作者:石滨萌(1993.02—)男,汉族,河北石家庄人,科研助理,硕士,研究方向:生物特征、图像处理方向。