Research And Application Of Vision-based Mold Protection Network Monitoring System

Injection molding machine is the main production equipment for plastic products.It is very important to ensure effective protection of molds in injection molding machines.The traditional manual inspectioan methods have problems such as high labor intensity,low detection efficiency,nd management confusion,which can no longer meet the requirements of industrial production.To solve the above problems,this paper researches and develops a system which based on machine vision and network.The system uses a visual method to detect the mold hanging mold,workpiece defects,through the port connects the mold protection terminal with the injection molding machine to read the production data,uses the network platform to monitor and manage the production process of injection molding machine.The main research work of this paper is as follows:1.An image matching algorithm based on the SURF feature descriptor is proposed as the image processing technology of the mold protection monitoring terminal.It has the good matching characteristics of constant scale and rotation invariant,which solves the problems caused by the offset of the mold position due to the vibration of the injection molding machine in the image matching process.It improves the accuracy of image detection.Image processing Uses Gaussian approximation filter and integral image method to improve the speed of image detection and ensure the real-time performance of the mold protection monitoring terminal.2.Developing a background monitoring center based on the.NET platform,using the TCP/IP transmission protocol,Socket communication to achieve the data exchange function between the background monitoring center and the field monitoring terminal;combining with multi-threading technology makes it possible to monitor multiple terminals at the same time;using Entity Framework database technology to achieve the background monitoring center for data storage and management.3.In order to reduce the coupling of user pages and back-end code,improve the scalability and maintainability of the client’s website,this article uses the ASP.NET MVC model to develop the client’s website and implements data query and remote monitoring functions.Through jQuery and Ajax technology to improve the update speed of the page information;using the Razor view engine to improve the security of the site;using the BootStrap front-end framework to layout and render the pages,making user pages more humanize.Compared with the traditional manual inspection method,the false detection rate of the vision-based mold protection network monitoring system is less than 2.5%,which meets the requirements of the enterprise,and enables highly efficient and concise networked production management,It has a high application value in the injection molding automatic production monitoring industry.