快速监督学习在显示器配件分类及识别中的应用

吴海波; 崔禹; 王森; 王晨; 潘云龙

doi:10.13433/j.cnki.1003-8728.20200402

快速监督学习在显示器配件分类及识别中的应用

doi: 10.13433/j.cnki.1003-8728.20200402

昆明理工大学机电工程学院，昆明　650500

基金项目: 国家自然科学基金地区基金（52065035）、云南省教育厅科学研究基金项目（2019J0045）及云南省级人培项目（KKSY201801018，KKSY201701001）

详细信息

作者简介:
吴海波（1975−），高级工程师，硕士生导师，博士，研究方向为机电产品设计与开发、机器人理论研究与应用，whb_kust@kust.edu.cn

通讯作者:
王森，讲师，硕士生导师，博士，wangsen0401@126.com

中图分类号: TG156
计量
- 文章访问数: 60
- HTML全文浏览量: 14
- PDF下载量: 9
- 被引次数: 0
出版历程
- 收稿日期: 2020-09-07
- 录用日期: 2021-12-17
- 刊出日期: 2022-09-05

Application of Fast Supervised Learning in Classification and Recognition of Display Accessories

Faculty of Mechanical and Electrical Engineering, Kunming University of Science and Technology, Kunming 650500, China

摘要

摘要: 针对常见显示器配件难以快速实时分类识别的问题，以显示器配件图像分类识别为核心，构建一种基于监督学习的显示器配件快速视觉识别系统。通过对生产线上实时采集的显示器配件图像进行低通中值滤波消除图像中的噪声、孤立亮点或暗点，使用高斯算子滤波削弱图像像素灰度变化，使图像表面均匀平滑；使用样本集对监督学习分类器进行6次训练；利用监督学习分类器对显示器配件进行分类识别。基于4种分类识别方法的实验对比结果表明：本文方法采用图像的预处理弥补了监督学习分类器因噪声影响而导致分类识别精度下降的不足，在实时性和鲁棒性方面明显优于其他3种分类识别方法，完成分类识别仅需12.9 ms，每一种配件的识别准确率达到96%以上，分类准确率达到100%，该算法满足显示器配件分类识别的工程应用及实时分拣需求。
- 分类识别 /
- 中值滤波 /
- 高斯算子 /
- 计算机视觉 /
- K-邻近法
Abstract: Aiming at the problem that common display accessories are difficult to quickly classify and recognize in real time, the image classification and recognition of display accessoriesis takenas the core, and a fast visual recognition system for display accessories based on the supervised learningis built. Firstly, low-pass median filtering is performed on the display accessories images collected in real time on the production line to eliminate the noise, isolated bright spots or dark spots in the image, and Gaussian filter is used to weaken the gray changes of image pixels to make the image surface uniform and smooth. Secondly, the sample set to train the supervised learning classifier for 6 timesis used. Finally, the supervised learning classifier to classify and recognize the display accessoriesis used. The experimental comparison results based on the four classification and recognition methods show that the present method uses the image preprocessing to make up for the insufficient classification andrecognition accuracy of the supervised learning classifier due to the influence of noise. It is significantly better than the other three in terms of the real-time and robustness. It takes only 12.9 ms to complete the classification and recognition, and the recognition accuracy of each accessory is over 96%, and the classification accuracy is 100%. The present algorithm meets the engineering application and real-time sorting requirements of display accessory classification and recognition.
- classification recognition /
- median filtering /
- gaussian operator /
- computer vision /
- K-proximity method

HTML全文

图 1 配件分类识别框架图

下载: 全尺寸图片幻灯片

图 2 二维高斯滤波器三维特性图

下载: 全尺寸图片幻灯片

图 3 分类识别算法检测平台

下载: 全尺寸图片幻灯片

图 4 4种不同算法的定性比较

下载: 全尺寸图片幻灯片

图 5 显示器配件识别准确率

下载: 全尺寸图片幻灯片

图 6 显示器配件识别准确率误差

下载: 全尺寸图片幻灯片

表 1 显示器配件预处理

采集图像	低通中值滤波	高斯滤波
1
2
3

下载: 导出CSV

表 2 配件类标准差

样本类别	支柱	底座	线材
样本数量	50	50	50
平均标准差	0.02	0.01	0

下载: 导出CSV

表 3 配件类距离表

样本类别	底座	线材	支柱
底座	0	1.68	1.74
线材	1.68	0	1.66
支柱	1.74	1.66	0

下载: 导出CSV

表 4 显示器配件分类准确率

识别算法	支柱分类准确率/%	线材分类准确率/%	底座分类准确率/%
本文算法	100	100	100
最小平均距离	89.44	91.82	93.75
KNN（灰度）	100	69.08	100

下载: 导出CSV

参考文献(23)

[1]	郭斐, 靳伍银, 王猛. 基于改进的Faster R-CNN算法的机械零件图像识别[J]. 机械设计, 2019, 36(9): 113-116 GUO F, JIN W Y, WANG M. Image recognition of mechanical parts based on the improved Faster R-CNN algorithm[J]. Journal of Machine Design, 2019, 36(9): 113-116 (in Chinese)
[2]	汪嘉杰, 王磊, 范秀敏, 等. 基于视觉的航天电连接器的智能识别与装配引导[J]. 计算机集成制造系统, 2017, 23(11): 2423-2430 WANG J J, WANG L, FAN X M, et al. Vision based intelligent recognition and assembly guidance of aerospace electrical connectors[J]. Computer Integrated Manufacturing Systems, 2017, 23(11): 2423-2430 (in Chinese)
[3]	晏开华, 苏真伟, 黄明飞. 支持向量机在机械零件识别中的应用[J]. 电子技术应用, 2008, 34(11): 108-110,114 YAN K H, SU Z W, HUANG M F. Application of SVM in recognition of mechanical parts[J]. Application of Electronic Technique, 2008, 34(11): 108-110,114 (in Chinese)
[4]	逄增治, 史建杰, 尹建芹, 等. 基于目标形态特征的工件自动分割方法[J]. 北京邮电大学学报, 2019, 42(5): 119-126 PANG Z Z, SHI J J, YIN J Q, et al. Automatic segmentation of workpiece based on target morphological features[J]. Journal of Beijing University of Posts and Telecommunications, 2019, 42(5): 119-126 (in Chinese)
[5]	金鹏, 刘检华, 刘少丽, 等. 基于二维靶标的管路端点位置测量方法[J]. 计算机集成制造系统, 2014, 20(11): 2758-2766 JIN P, LIU J H, LIU S L, et al. Measuring method of pipeline endpoints based on two-dimensional point-target[J]. Computer Integrated Manufacturing Systems, 2014, 20(11): 2758-2766 (in Chinese)
[6]	王立忠, 赵建博, 谈杰, 等. 高强钢薄板高温焊接变形的视觉测量[J]. 光学精密工程, 2020, 28(2): 283-295 WANG L Z, ZHAO J B, TAN J, et al. Visual measurement of high-temperature welding deformation for high-strength steel sheet[J]. Optics and Precision Engineering, 2020, 28(2): 283-295 (in Chinese)
[7]	孙丽萍, 陈果, 陈立波, 等. 基于KPCA的航空发动机滑油滤磨屑图像识别[J]. 机械科学与技术, 2010, 29(6): 731-736 SUN L P, CHEN G, CHEN L B, et al. Image recognition of aero-engine oil filter debris by kernel principle component analysis[J]. Mechanical Science and Technology for Aerospace Engineering, 2010, 29(6): 731-736 (in Chinese)
[8]	陈付梦, 王静秋, 张龙. 基于提升小波的铁谱图像边缘检测[J]. 机械科学与技术, 2013, 32(10): 1466-1470 CHEN F M, WANG J Q, ZHANG L. Edge detection of ferrography image based on lifting wavelet[J]. Mechanical Science and Technology for Aerospace Engineering, 2013, 32(10): 1466-1470 (in Chinese)
[9]	CAO S Q, YANG G L, ZHU Q Y, et al. A novel feature extraction method for mechanical part recognition[J]. Applied Mechanics and Materials, 2011, 88-89: 116-121 doi: 10.4028/www.scientific.net/AMM.88-89.116
[10]	葛为民, 申耀华, 王肖锋. 箱梁结构件焊缝表面缺陷特征提取及分类研究[J]. 仪器仪表学报, 2018, 39(12): 207-215 GE W M, SHEN Y H, WANG X F. Feature extraction and classification of weld surface defects in box girder structures[J]. Chinese Journal of Scientific Instrument, 2018, 39(12): 207-215 (in Chinese)
[11]	谢长贵, 谢志江. 热态重轨表面缺陷机器视觉检测的关键技术[J]. 重庆大学学报, 2013, 36(10): 16-21 XIE C G, XIE Z J. Key technology of detecting hot heavy rail steel surface faults based on machine vision[J]. Journal of Chongqing University, 2013, 36(10): 16-21 (in Chinese)
[12]	李静蕊, 王刚, 周运金, 等. 基于ART2神经网络的机械零件模式识别[J]. 哈尔滨工业大学学报, 2009, 41(3): 117-120 LI J R, WANG G, ZHOU Y J, et al. Workpiece pattern recognition based on ART2 nerual network[J]. Journal of Harbin Institute of Technology, 2009, 41(3): 117-120 (in Chinese)
[13]	李养胜, 李俊. 基于支持向量机与k-近邻的工件表面缺陷识别算法[J]. 电子测量技术, 2018, 41(7): 50-53 LI Y S, LI J. Workpiece surface defect recognition algorithm based on SVM and k-nearest neighbor[J]. Electronic Measurement Technology, 2018, 41(7): 50-53 (in Chinese)
[14]	刘桂华, 王玉玫, 王静强. 基于机器视觉的隔水管法兰端面位姿检测[J]. 机械设计, 2019, 36(3): 84-90 LIU G H, WANG Y M, WANG J Q. Riser flange flat pose detection based on the machine vision[J]. Journal of Machine Design, 2019, 36(3): 84-90 (in Chinese)
[15]	吴翰. 数字图像的高斯噪声去噪算法研究 [D]. 安庆: 安庆师范大学, 2018 WU H. Research on denoising algorithm of gauss noise in digital image [D]. Anqing: Anqing Normal University, 2018 (in Chinese)
[16]	赵高长, 张磊, 武风波. 改进的中值滤波算法在图像去噪中的应用[J]. 应用光学, 2011, 32(4): 678-682 ZHAO G C, ZHANG L, WU F B. Application of improved median filtering algorithm to image de-noising[J]. Journal of Applied Optics, 2011, 32(4): 678-682 (in Chinese)
[17]	王永明, 王贵锦. 图像局部不变性特征与描述 [M]. 北京: 国防工业出版社, 2010 WANG Y M, WANG G J. Image local invariant features and descriptors [M]. Beijing: National Defense Industry Press, 2010 (in Chinese)
[18]	COVER T, HART P. Nearest neighbor pattern classification[J]. IEEE Transactions on Information Theory, 1967, 13(1): 21-27 doi: 10.1109/TIT.1967.1053964
[19]	萨日娜, 胡志勇, 张秀芬, 等. 面向维修性设计的复杂产品设计方案评价方法[J]. 机械设计与制造, 2014(11): 61-63,67 SA R N, HU Z Y, ZHANG X F, et al. Complex product design scheme evaluation method for maintainability design[J]. Machinery Design & Manufacture, 2014(11): 61-63,67 (in Chinese)
[20]	DENG Z Y, ZHU X S, CHENG D B, et al. Efficient kNN classification algorithm for big data[J]. Neurocomputing, 2016, 195: 143-148 doi: 10.1016/j.neucom.2015.08.112
[21]	ZHANG S C, LI X L, ZONG M, et al. Learning k for kNN Classification[J]. ACM Transactions on Intelligent Systems and Technology, 2017, 8(3): 43
[22]	XIE H H, LIANG D, ZHANG Z J, et al. A novel pre-classification based kNN algorithm[C]//2016 IEEE 16th International Conference on Data Mining Workshops. Barcelona: IEEE, 2016: 1269-1275
[23]	MARTÍNEZ-NÚÑEZ M, PÉREZ-AGUIAR W S. Efficiency analysis of information technology and online social networks management: an integrated DEA-model assessment[J]. Information & Management, 2014, 51(6): 712-725