基于ARM的陶瓷划痕自动检测装置设计研究

【作者】 王健；

【导师】 张重雄；

【作者基本信息】 南京理工大学 ， 通信与信息系统， 2009， 硕士

【摘要】 本文分析了国内外陶瓷划痕检测的发展和现状,介绍了一种嵌入式陶瓷划痕自动检测装置。该检测装置设计以ARM处理器为核心,先通过CMOS摄像头采集陶瓷图像,将其存储在ARM存储器里,再经过图像预处理,边缘检测,图像平滑,二值化处理和划痕识别,判定陶瓷是否有划痕,并将检测数据显示出来,如果有划痕,将有划痕的陶瓷图片通过以太网从该装置传送到主控计算机。论文首先介绍了陶瓷划痕检测装置的硬件架构和软件平台的构建,其次重点介绍了陶瓷划痕检测的数字图像处理技术,对其中的一些算法进行了详细地分析,并在Matlab下进行了仿真,同时对用于图像边缘检测的Prewitt算子进行了算法改进,减少了图像处理的运算量,更好地满足了该检测装置的实时性要求。最后详细介绍了该装置的软件设计,这也是本论文最核心的一部分,软件设计包括陶瓷图像的采集,划痕的检测,按键模块,显示模块和通讯模块等。本文的研究结果具有较大的工程实际意义,对于陶瓷划痕的检测具有一定的参考价值。更多还原

【Abstract】 By analyzing the development and situations of ceramic-nick inspecting system home and abroad, this thesis shows an embedded automatic ceramic-nick inspecting system. The system based on ARM microprocessor, and it gets image data which will be stored in ARM with a CMOS camera, and then carries out a series of processing, including image preprocessing, edge detection, image smoothing, image binaryzation, automatic recognition. Finally, it detects the ceramic whether has nick or not by calculating some parameters which will be showed. If the ceramic has nick, the image will be transported from the device to main controlling computer.Firstly, the hardware design and the set up of software platform are put forward and discussed. Secondly, the thesis makes a study of digital image processing technology. Some important algorithms has been analyzed in detail, and simulated with Matlab, at the same time, algorithm of Prewitt functor which is used in edge detection has been improved to reduce calculated amount of image processing and has satisfied the demand of Real-Time of the detecting device. Finally, this thesis discussed the software design of system in detail, which is hardcore of the thesis, including capturing image, inspecting ceramic-crack, the design of buttons and display, communication module and so on.Result of this research was provided with great signality of practical engineering and a valuable reference for ceramic-nick inspecting.更多还原