Development Of Defects Detection Client Software For Digital Textile Printing Based On Embedded GPU

Digital printing technology is the crux technology to promote the transformation and upgrading of printing and dyeing industry in China.Digital printing machine developed in domestic mostly depend on artificial detection method to realize the quality detection of the products at present,which result in low detection accuracy and high labor costs.With the development of machine vision technology,the automatic detection and identification of printing defects occurred in the process of digital printing has become industry development trend.The digital printing defect detecting system studied in this thesis is based on actual digital printing machine platform,it configures a real-time data processing unit and an automatic image detection unit based on embedded GPU,designs the GUI client software,achieves the control of the digital printing process and automatic detection,which is of great value in application.The defect detection client software designed in this thesis is an important part of the digital printing system,which controls and manages the whole system through the interaction with users,and improves the convenience and versatility of the system.The client software bases on Qt graphics development framework and runs at Windows 7 operating system.This thesis designs the structure of the client software,divide it into interface layer and logical business layer,divides it into configuration,image detect,users management,warning,history and processor modules based on modular thought.It achieves the management of mass data based on SQLite database,designs SDK interface and communication protocol to communicate with server.It designs image display module through QGraphicsView structure of Qt,which has functions of image magnification,narrow,translation and area selection.It achieves the query and management of history defective images based on NTFS file management system.It achieved a method based on LBP to detect defective images.Through strict tests,this client software has process control and automatic image detection and display functions,which can satisfy the requirements of the system.