基于软件工程与多模块的工业图像在线处理系统

INDUSTRIAL IMAGE ONLINE PROCESSING SYSTEM BASED ON SOFTWARE ENGINEERING AND MULTI-MODULE

为了提高工业图像处理效率,通过开发一站式软件系统,设计了一种基于软件工程与多模块的工业图像在线处理系统。首先,开发一个基于图像帧文件联合存储的图像制式合并模块,将数百图像帧合并为一个便于传输和管理的大文件。然后,设计一种基于人机交互的图像浏览模块,实现图像特征标注与浏览。随后,结合线程机制与百度云OSS,开发上传模块,达到远程存储与分析的目的,并将图像分析模块与图文报告模块融合,自动给出工业图像分析结果。最后,基于QTcreator开发平台与QT语言编程来实现各个模块的功能。测试结果显示:与单一模块的图像处理系统相比,所提系统具有更好的在线处理性能,具备更高的作业稳定时间。

In order to improve the efficiency of industrial image processing process, this paper designs an industrial image online processing system based on software engineering and multi-module which developing one-stop software system. Firstly, an image standard merging module based on joint storage of image frame files is developed to merge hundreds of image frames into a large file for easy transmission and management. Then, an image browsing module based on human-computer interaction is designed to achieve the goal of image feature annotation and browsing. Subsequently, combined with thread mechanism and Baidu Cloud OSS, the upload module is developed to achieve the purpose of remote storage and analysis. The image analysis module and the graphic report module are fused to automatically give the results of industrial image analysis and assist the industrial quality analysis department in product diagnosis. Finally, based on the development platform of QTcreator and the programming of QT language, each module is realized and the complete system is constructed. The test results show that the proposed system has better on-line processing performance and higher job stabilization time compared with the single module image processing system.