发布时间 : 星期五 文章图像阈值分割算法分析与实现更新完毕开始阅读e4e339ea524de518964b7d44
九江学院
JIU JIANG UNIVERSITY
毕 业 论 文（设 计）
题 目 图像阈值分割算法分析与实现 英文题目 The analysis and implementation of image
threshold segmentation algorithms 院 系 电子工程学院 专 业 通信工程 姓 名 赖小春 年 级 2008级 指导教师 谭小容
二零一二年六月
九江学院学士学位论文
摘 要
图像阈值分割算法分析与实现是图像处理领域的一个基本的、重要的研究课题。所谓图像分割是指根据灰度、彩色、空间纹理、几何形状等特征把图像划分成若干个互不相交的区域，使得这些特征在同一区域内，表现出一致性或相似性，而在不同区域间表现出明显的不同。
图像阈值分割法是一种最常用，同时也是最简单的图像分割方法，它特别适用于目标和背景占据不同灰度级范围的图像。它不仅可以极大的压缩数据量，而且也大大简化了分析和处理步骤。因此在很多情况下，是进行图像分析、特征提取与模式识别之前的必要的图像预处理过程。图像阈值化的目的是要按照灰度级对像素集合进行一个划分，得到的每个子集形成一个与现实景物相对应的区域，各个区域内部具有一致的属性，而相邻区域布局有这种一致属性。这样的划分可以通过从灰度级出发选取一个或多个阈值来实现。本文首先介绍了图像分割发展现状，其次对图像分割的基础做了简单的介绍，最后重点对双峰法阈值分割、分水岭阈值分割、0tsu阈值分割作了详细分析与研究，并且把这三种算法的分割效果进行了简单的比较，结果发现各阈值分割方法都有各自的优劣性，需要根据图像的实际情况选择适合的方法。分割结果的好坏或者正确与否，目前还没有一个统一的评价判断准则，分割的好坏必须从分割的效果和实际应用场景来判断。
关键词：直方图；图像分割；阈值；算法
I
图像阈值分割算法分析与研究
The analysis and implementation of image threshold
Segmentation algorithm
Abstract
The analysis and implementation of image threshold segmentation algorithm is a
basic and important research topic. The so-called image threshold segmentation is mean to divide the grayscale, color, space, texture, geometry and other features into several disjoint areas, so that these characteristics in the same area, showing the consistency or similarity, while showing different in different regions .
The segmentation of image threshold method is one of the most commonly and simple image segmentation method, it is especially suitable for the target and background occupy different gray-scale range of the image. It not only can greatly compress the amount of data, but also greatly simplifies the analysis and processing steps. In many cases, image analysis, feature extraction and pattern recognition are till the necessary image pre-processing process. The purpose of image threshold is to divide each subset of the formation of a region corresponding into the realistic scenery according to the gray level pixel, each region consistent with the properties of adjacent regional distribution of this a consistent attribute. This division can select one or more threshold starting from the gray level. This paper first introduces the development status of image segmentation, followed by a brief introduction on the basis of image segmentation, and finally focus on the apex method threshold segmentation, watershed threshold segmentation and 0tsu threshold segmentation are analyzed and researched in detail, and made a simple comparison with the segmentation results of the three algorithms and found that each threshold segmentation method has its own advantages and disadvantages, you need to select the appropriate method according to the actual situation of the image. There are no single evaluation criterion about Segmentation result is good or bad, right or wrong, , the segmentation good or bad are based on the segmentation results and determine on the actual scenarios .
Keywords: histogram; image segmentation; threshold; algorithm
II