MeanShift图像分割方法.doc

摘 要 在图像处理和计算机视觉里，图像分割是一个十分基础而且很重要的部分，决定了最终分析结果的好坏。图像分割问题的典型定义就是如何在图像处理过程中将图像中的一致性区域和感兴趣对象提取出来。 MeanShift 图像分割方法是一种统计迭代的核密度估计方法。MeanShift算法以其简单有效而被广泛应用，但该方法在多特征组合方面和数据量较大的图像处理上仍存在不足之处，本文针对这些问题对该算法的结构进行了优化。本文利用图像上下文信息对图像进行了区域合并以此来对输入数据进行了压缩；并实现特征空间中所有特征量的优化组合。 最后，总结了本文的研究成果。下一步需要深入的研究工作有:(1)考虑分割的多尺度性,实现基于Mean Shift算法的多尺度遥感图像分割;(2)考虑利用Gabor滤波器来提取纹理特征,或将更多的特征如形状等特征用于MeanShift遥感图像分割中。 关键词: Mean Shift, 图像分割, 遥感图像, 带宽 ABSTRACT mage segmentation is very essential and critical to image processing and computer vision, which is one of the most difficult tasks in image processing, and determines the quality of the final result of analysis. In image segmentation problem, the typical goal is to extract continuous regions and interest objects in the case of image processing. The Mean Shift algorithm for segmentation is a statistical iterative algorithm based on kernel density estimation. Mean Shift algorithm has been widely applied for its simplicity and efficiency. But the algorithm has some deficiencies in feature combination and image processing for large data. According to the deficiencies of the Mean Shift algorithm, this paper optimizes the structure of the algorithm for segmentation. Firstly, this paper introduces a method of data compressing by merging the nearest points with similar properties into consistency regions. Secondly, We optimize the combination of features. At last, after concluding all research work in this paper, further work need to be in-depth studied: (1) Consider multi-scale factors of remote sensing, and realize multi-scale remote sensing image segmentation based on Mean Shift algorithm. (2) Consider extracting textures features by using Gabor filter, or use more features such as shape features to segment remote sensing images based on Mean Shift algorithm. KEY WORDS: Mean Shift, image segmentation, remote sensing images, bandwidth, 目 录 第一章 绪 论 1 1.1选题背景 1 1.2国内外研究现状及进展 2 1.2.1 Mean Shift算法的国内外研究现状 2 1.2现有研究存在的问题