基于数学形态学的图像分割算法.doc

基于数学形态学的图像分割算法 摘 要 在过去的近二十年里,图像分割作为计算机视觉等领域的重要研究方向,吸引了越来越多的研究者的注意，由于形态学技术在进行图像处理时，充分考虑到了图像的结构特征，因此相对于其他图像处理方法而言，形态学技术具有独特的结构特征优势。 本文以数学形态学为基础研究了图像处理算法。从数学形态学的基本思想和基本理论入手，将从数学形态学在图像滤波、边缘检测以及图像分割方面的三个应用作一一介绍，同时，将形态学图像处理方法与传统图像处理方法进行了比较。 ①本文利用腐蚀、膨胀、开闭运算等基本的形态学运算及其组合对图像进行滤波，同时将形态学滤波方法与传统滤波方法效果进行比较。 ②文章介绍了基本的形态学边缘检测方法，如内边界、外边界、Top-Hot变换等，针对本文要处理的卟啉阵列图像，提出了一种改进的BM方法，取得了较好的效果，最后将形态学边缘检测与传统边缘检测方法进行了比较。 ③本文主要采用数学形态学的分水岭算法对图像进行分割，详细论述了分水岭算法的原理及其算法步骤，由于图像中噪声以及细小边缘的存在，分水岭算法很容易产生过分割现象，我们采取了两种方法减弱过分割现象：先利用形态学滤波器进行预处理，然后再利用区域生长型分水岭算法对图像进行分割。研究了几种常见的分割方法，将其分割效果与分水岭算法的分割效果进行了比较。最后,对标准LENA图像进行了形态学方法处理. 关键词：数学形态学,滤波,边缘检测,分水岭  ABSTRACT In past almost twenty years, image segmentation, as an important research field in computer vision, attracts the attention of more and more researchers. Compared to other image processing methods, the mathematical morphology method takes full account of the structural characteristics of the image, so it has the unique advantage in image processing. In this paper, image processing algorithms are researched based on mathematical morphology. We analysis the applications of mathematical morphology in image filtering, edge detection and image segmentation separately, at the same time, compare traditional image processing methods with mathematical methods. This paper uses basic Morphological methods, erosion, dilate, opening，closing, and thecombinations to filt. classical Filting methods are compared with Morphological Filting in image processing. ② This paper introduces basic Morphological gradients，such as the inner boundary, the uter boundary and Top-hot Transform, as well as an improved BM methods, which made a good result when applied to detect porphyrin array images. At last, classical gradients are compared with Morphological gradient. ③ In this paper, the watershed algorithm of mathematical morphology is used for image segmentation. We discuss the principle and steps of watershed algorithm in image processing. The watershed algorithm is easy t