基于OpenCV与USB工业相机零件检测系统的图像处理研究  

Image processing of detection system for parts in industrial camera based on OpenCV and USB

作　　者：黄振峰[1] 陈海平[1] 邓培[1] 张鑫星[1] 

机构地区：[1]广西大学机械工程学院,广西南宁530004

出　　处：《现代电子技术》2012年第18期128-132,共5页Modern Electronics Technique

基　　金：国家科技重大专项资助项目(2008ZX07317-001-002)

摘　　要：针对OpenCV库函数在Linux平台下不能直接获取非V4L接口的USB工业相机视频流和IplImage结构的图像不能直接显示于Qt控件中的问题,提出了一种新的解决方法:共享数据流法,并在Qt集成开发环境下完成共享数据流法的程序代码编写;最后通过与原有方案的实验对比,验证了共享数据流法的可行性和优越性,为实现高速高精度机械零件检测系统打下良好的基础。A novel approach, named shared-streams, is proposed in this paper to tackle the two problems that OpenCV library function can＇t directly obtain video data flow from the USB industrial cameras with non-V4L interface on the Linux plat- form and the Ipllmage structure images can＇t immediately be showed in Qt control. The code of the shared-streams solution is written under the Qt integrated developing environment. In experimental comparison with the traditional program, the results indicate the feasibility and the advantages of shared-streams approach, which has established a solid foundation for the mechanical part detection system with high speed and precision.

关 键 词：图像处理 OPENCV USB工业相机 LINUX 

分 类 号：TN919-34[电子电信—通信与信息系统] TP319[电子电信—信息与通信工程]