数字图像的边缘检测毕业设计论文.doc

内蒙古科技大学毕业设计说明书（毕业论文） PAGE II 摘 要 数字图像的轮廓提取是数字图像处理中的一个重要方面，目的是在一幅图像中提取对象的外部轮廓，为下一步的形状分析和目标识别做准备，是许多有关图像研究的重要中间环节。 本文讲述了怎样对数字图像进行简单的轮廓提取, 包括图像的预处理过程和轮廓提取的一般方法。首先要对原始图像进行预处理，处理方法主要有图像平滑、中值滤波、梯度锐化和拉普拉斯锐化等。图像平滑处理能去除部分噪声干扰，但同时也弱化了图像的边缘轮廓，而中值滤波处理去除噪声的效果相对来说要好。梯度锐化、拉普拉斯锐化等处理方法能增强图像的边缘轮廓，也相对的弱化了噪声对轮廓提取的干扰。通过对图像进行一些预处理后，就可以用不同的边缘检测算法检测出不同图像的边缘轮廓，然后通过软件实现掏空目标区域的内部点，来提取出图像的轮廓。 从图像中提取出来的轮廓可以用于进一步的图像识别、数学特征计算等研究。它的应用是很广泛的，比如医学图像、电影、电视、出版物、摄影等等。 本文是基于Visual C++环境实现的数字图像的轮廓提取，所用编程工具为Visual Studio 2005，它能帮助用户直观的、可视地设计程序的用户界面，可以方便的编写和管理各种类，维护程序源代码，因此能有效的提高开发效率。 关键词:图像预处理；边缘检测；轮廓提取 Abstract Digital image processing contour extraction is an important aspect in Digital Image Processing.It’s purpose is to extract objects in an image of the external contour. Preparing for the next shape analysis and object recognition. It’s the image of many of the important intermediate links. This article describes how to extract the contour of digital images simply. Including image pre-processing process,and the general contour extraction method. First of all is the original image pre-processing. Treatment methods have Image smoothing, median filtering, gradient sharpening and Laplacian sharpening etc. Image smoothing could remove part of the noise, But it also weakened the image edges,and median filtering can remove noise better. Gradient sharpening and Laplacian sharpening can enhance the image edges, they also have a relative weakening of the interference noise on the contour extraction.After some pre-processings to the digital image, we can use different edge detection algorithms to detect different egde of images,then empting the internal point of the target area to extract the contour of images. Contours extracted from the feature can be used for further image recognition, mathematical calculation of characteristic. Its application is very broad, such as medical images, movies, television, publications, photography, etc. This article is based on Visual C++ implementation of the digital image