基于模式分类的喷嘴料液雾滴速度测量研究

doi:10.16180/j.cnki.issn1007-7820.2019.08.007

摘要/Abstract

摘要：

测量烟叶加料过程中料液雾滴喷射后的速度变化,为烟叶加料均匀性的评价提供有效数据。实验从喷嘴的轴向面采集喷出的雾滴图像,对采集到的图像进行预处理和图像分割,采用模式分类对分割出的雾滴分类,根据高速相机成像的测量方法计算出每一类雾滴的速度。研究结果表明,由模式分类统计出的数据,其准确率达到95.65%,使用该数据计算出的速度误差为4.6%,两项数据均在规定范围之内。使用高速摄像机获取图像,采用软件计算出雾滴速度,与早期实验方案相比,降低了测量设备价格,实现低成本测量;模式分类的测量方法准确率较高,为后续加料均匀性分析提供可靠的数据,且效率更高,同时为数据测量的智能化打下基础。

关键词: 雾滴速度, 图像预处理, 图像分割, 模式分类, 低成本, 智能化

Abstract:

In order to measure the change of velocity after the spray of liquid droplets in the process of tobacco leaf feeding, it provided effective data for the evaluation of the uniformity of tobacco feed. In this experiment, the image of the sprayed droplets were collected from the axial direction of the nozzle, and the collected images were pre-processed and image segmentation. Classifying the segmented droplets by pattern classification, and calculating the velocity of each type of droplet according to the measurement method of high-speed camera imaging. The results shown that the accuracy of the data calculated by the model classification was 95.65%, and the calculated speed error was 4.6%. Both data were within the specified range. The experiment used a high-speed camera to acquire images, and the droplet velocity was calculated by software. Compared with the earlier experimental scheme, reduced measurement equipment prices for low-cost measurements. The measurement method of the pattern classification had higher accuracy, which could provide reliable analysis for subsequent feeding uniformity analysis. The data was more efficient and at the same time played the foundation for the intelligence of data measurement.

Key words: droplet velocity, image preprocessing, image segmentation, pattern classification, low-cost, intelligentialize

中图分类号:

TP391.41

孙腾飞,蒋红海. 基于模式分类的喷嘴料液雾滴速度测量研究[J]. 电子科技, 2019, 32(8): 33-37.

SUN Tengfei,JIANG Honghai. Research on Speed Measurement of Nozzle Material Liquid Droplet Based on Pattern Classification[J]. Electronic Science and Technology, 2019, 32(8): 33-37.

图/表 8

图1

图2

图3

图4

图5

图6

表1

表2

参考文献 19

[1]	蒋汉城 . 浅析烟草制丝工艺过程中可提高的几个方面[J]. 中国化工贸易, 2017,9(14):80-80.
	Jiang Hancheng . Analysis of several aspects that can be improved in the process of tobacco spinning[J]. China Chemical Trade, 2017,9(14):80-80.
[2]	陈良元 . 卷烟生产工艺技术[M]. 郑州: 河南科学技术出版社, 2002.
	Chen Liangyuan. Cigarette production technology[M]. Zhengzhou: Henan Science and Technology Press, 2002.
[3]	曹建明 . 喷雾学[M]. 北京: 机械工业出版社, 2005.
	Cao Jianming. Spray science[M]. Beijing: Machinery Industry Press, 2005.
[4]	闻志勇 . 风幕式喷杆喷雾雾滴运动特性与飘移性能试验研究[D]. 镇江:江苏大学, 2016.
	Wen Zhiyong . Experimental investigation of droplet characteristics and drift performance in air-assist boom sprayer[D]. Zhenjiang:Jiangsu University, 2016.
[5]	刘海龙, 陈孝震, 蔡小舒 , 等. 基于轨迹图像的气液旋风分离器液滴粒度、浓度、速度的在线测量[J]. 化工学报, 2012,63(6):1729-1734. doi: 10.3969/j.issn.0438-1157.2012.06.011
	Liu Hailong, Chen Xiaozhen, Cai Xiaoshu , et al. In-line measurement of size,concentration and velocity of drops from gas-liquid cyclone separator based on trajectory image processing[J]. CIESC Journal, 2012,63(6):1729-1734. doi: 10.3969/j.issn.0438-1157.2012.06.011
[6]	梁坤峰, 彭正标, 袁竹林 , 等. 液液雾化特性与粒径分布规律[J]. 化工学报, 2007,58(8):1935-1942.
	Liang Kunfeng, Peng Zhengbiao, Yuan Zhulin , et al. Atomization and drop-size distribution of liquid-liquid systems[J]. CIESC Journal, 2007,58(8):1935-1942.
[7]	史春建, 邱白晶, 汤伯敏 , 等. 基于高速图像的雾滴尺寸分布统计与运动分析[J]. 农业机械学报, 2006,37(5):63-66.
	Shi Chunjian, Qiu Baijing, Tang Bomin , et al. Analysis of droplet size distribution and movement based on high speed image processing[J]. Transactions of the Chinese Society for Agricultural Machinery, 2006,37(5):63-66.
[8]	吴学成 . 基于轨迹图像的煤粉颗粒速度和粒径测量[J]. 浙江大学学报:工学版, 2011,45(8):136-140.
	Wu Xuecheng, Wang Huai, Hu Qian , et al. Velocity and size measurement of coal particle by trajectory imaging[J]. Journal of Zhejiang University:Engineering Science, 2011,45(8):136-140.
[9]	杜永成, 杨立, 刘天 , 等. 基于数字粒子图像测速的水雾粒径测量算法及实验[J]. 中国光学, 2013,6(3):409-414.
	Du Yongcheng, Yang Li, Liu Tian . Algorithm and experiment on droplet sizing based on DPIV[J]. Chinese Optics, 2013,6(3):409-414.
[10]	解博, 周律 . 基于机器视觉的杏核图像识别系统设计[J]. 电子科技, 2016,29(10):97-100.
	Xie Bo, Zhou Lv . Design of an almond image recognition system based on machine vision[J]. Electronic Science and Technology, 2016,29(10):97-100.
[11]	洪西洋, 杨晖, 李然 , 等. 基于图像处理的颗粒流动层模型研究[J]. 电子科技, 2016,29(10):89-92.
	Hong Xiyang, Yang Hui, Li Ran , et al. Flow model based on image processing[J]. Electronic Science and Technology, 2016,29(10):89-92.
[12]	Carsten Steger, Markus Ulrich, Christian Wiedemann . 机器视觉算法与应用[M].杨少荣, 译. 北京: 清华大学出版社, 2008.
	Carsten Steger, Markus Ulrich, Christian Wiedemann. Machine vision algorithms and applications[M].Yang Shaorong, Translation. Beijing: Tsinghua University Press, 2008.
[13]	Pan Zhibin, Wu Xiuquan, Li Zhengyi , et al. Local adaptive binary patterns using diamond sampling structure for texture classification[J]. IEEE Signal Processing Letters, 2017,6(24):828-832.
[14]	Smołka P, Czerkawski M . The color wheel: a heuristic aid in classification of a skeletal pattern[J]. International Journal of Orthodontics, 2016,27(1):57-60.
[15]	Ur Rehman Sadaqat, Tu Shanshan, Huang Yongfeng , et al. A novel unsupervised convolution neural network approach for visual pattern classification[J]. AI Communications, 2017,30(5):311-324.
[16]	Adel Hafiane, Kannappan Palaniappan, Guna Seetharaman . Joint adaptive median binary patterns for texture classification[J]. Pattern Recognition, 2015,48(8):2609-2620.
[17]	刘渭滨, 邹智元, 邢薇薇 . 模式分类中的特征融合方法[J]. 北京邮电大学学报, 2017,40(4):1-8.
	Liu Weibin, Zou Zhiyuan, Xing Weiwei . Feature fusion methods in pattern classification[J]. Journal of Beijing University of Posts and Telecommunications, 2017,40(4):1-8.
[18]	薛梦霞, 刘士荣, 王坚 . 基于机器视觉的动态多目标识别[J]. 上海交通大学学报, 2017,51(6):727-733.
	Xue Mengxia, Liu Shirong, Wang Jian . Dynamic multi-target recognition based on machine vision[J]. Journal of Shanghai Jiaotong University, 2017,51(6):727-733.
[19]	张超 . 基于轨迹图像喷嘴液滴粒径和速度测量系统构建[J]. 烟草科技, 2014(2):8-11.
	Zhang Chao . Development of atomized particle size and velocity measurement system based on particle trajectory images[J]. Tobacco Science & Technology, 2014(2):8-11.

区域	A	B	D
传统统计	6.359 2	4.921 3	3.560 1
模式分类	6.104 8	4.704 7	3.396 3