摘要随着工业生产的发展，流水线生产与定制化生产逐渐成熟，对产品的检测要求随之提高。传统的人工检测方式无法适应高速的流水线生产，利用传感技术检测花费高昂。利用数字图像处理的检测方式逐渐成为产品检测的主流，它有着成本低廉，适用广泛，检测精度高、速度快、可靠性好等优势。为利用数字图像处理技术进行产品检测，对图像的处理方式和处理的具体算法是极为重要的一环，直接决定着产品检测的效率和精度。77163

本文着眼于产品检测的图像预处理算法选择和Hough变换进行图形拟合算法。分别从图像滤波、图像增强、图像分割、图像形态学处理和图形拟合五个模块设计产品检测程序。在Visual Studio 2015 平台上设计MFC应用程序进行仿真，以显微镜下纤维的直径和瓶盖的半径数值为检测目标。经实验，检测效果较好。

毕业论文关键词  产品检测  图像滤波  图像增强  图像分割  形态学处理  Hough变换

Title    Product testing based on digital image processing

Abstract  With the development of industrial production, production lines and customized production gradually matured, product testing requirements increase。 Traditional manual testing methods can not adapt to high-speed assembly line production。 Besides, product testing which uses the sensor technology is costly。 Digital image processing detection method gradually become a mainstream product detection technology, it has a low cost and wide applicability。 At the same time it has a high detection precision, speed and reliability。 For the use of digital image processing technology for product testing, specific algorithm for image handling and processing is a very important part which directly determines the efficiency and accuracy of product testing。

    This paper focuses on image preprocessing algorithm selection and Hough transform algorithm for pattern matching。The product testing program is designed from five parts which consists of image filtering, image enhancement, image segmentation, morphology processing and graphics fit。Simulate on the Visual Studio 2015 platform using MFC application。The test results are satisfactory。

Keywords  product testing  image filtering  image enhancement  image segmentation  morphology processing  Hough transform

目   次

1  引言 1

1。1  课题研究背景、目的和意义 1

1。2  论文内容提要 2

2  图像预处理算法研究 3

2。1  图像滤波 3

2。2  图像增强 7

2。3  图像二值化 10

3  图像的边缘检测与细化腐蚀 14

3。1  图像边缘检测算法与实现 14

3。2  图像腐蚀与细化算法与实现 18

4  图形拟合与参数测量 21

4。1  Hough变换原理 21

4。2  改进的Hough变换检测直线 25

5  产品检测软件系统构成与实现 28

5。1  产品检测处理流程