Research On Key Technology Of MES Based On Machine Vision Data Acquisition

Under the strategic plan of "Made in China 2025" and the urgent need of transformation and upgrading of the manufacturing industry in China,it has been a trend for the construction of digital factories.According to current production status of a domestic towel manufacturer and its transformation and upgrading needs,this paper carries out the research of the key technologr of MES based on machine vision data acquisition.This paper aims to achieve real-time and accurate collection of workshop production data and the transparent and Information management of production process of enterprise.The main research content is as follows.Firstly,according to the production characteristics and process flows of textile enterprises,this paper proposes the overall design scheme of MES by analyzing the production status of enterprises and the functional requirements of enterprise information management.Then key technologies needed for the implementation of the scheme are analyzed in detail.Secondly,the acquisition system of production data in enterprise manufacturing workshop based on machine vision technology is developed,this system includes character recognition module and curve recognition module.The character recognition module is developed through setting up a training sample,designing a classifier,pre-processing the image to be recognized and using the classifier.KNN algorithm and SVM algorithm are used to design the classifier of character recognition module respectively,the SVM algorithm is finally selected to design the character recognition module by comparing and analyzing the realization effect of these two algorithms.Thirdly,during the design of curve recognition module in data acquisition system,the curve image is pre-processed by rectifying the image,cutting out the curve area,converting the curve area image into the binary image,dilating the image and eroding the image.Then the recognition of curve image is achieved by calculating the coordinate transformation ratio factor and extracting curve data Finally,the two modules are integrated to achieve real-time and accurate collection of workshop production data.Lastly,SQL Server 2012 software is used to build MES database,the data acquired by the data acquisition system based on machine vision and the information about staffs and facilities got from ERP are saved into this database.The MES system based on this database is implemented by using C#programming language.The MES system mainly includes production progress module,production process monitoring module and equipment management module.