最小误差阈值分割方法研究+基于灰度图像的最小误差阈值法

Abstract
Image segmentation is a critical technique in image processing. Image thresholding is an image segmentation technique of simple computation and strong practicability. This article does research on several existing minimum error image thresholding methods for gray level images, then extends the idea of minimum error thresholding into color image segmentation and proposes 1D and 2D vector image thresholding methods, respectively. Firstly, 1D minimum error thresholding is studied, which has good time and space efficiency. However, it is sensitive to noise. In order to overcome this problem, researches tried to construct 2D histogram via integrating average gray level of local neighbourhood and extend 1D method to 2D one, thus improving noise immunity obviously. Unfortunately, the 2D thresholding method faces high computational complexity. Thus, a recursive idea is used to decrease computational complexity. In addition, as the 2D thresholding method can not accurately divide edge and noise region, object and background, researchers developed minimum error thresholding based on 2D histogram oblique segmentation. The above minimum error thresholding approaches can only deal with gray level images and do not consider color information. Therefore, with the motivation of other color thresholding approaches, we extend minimum error thresholding into color image segmentation, and develop two color image thresholding methods based on 1D and 2D vectors, respectively. The two color methods are useful attempts for the application of minimum error thresholding in color image segmentation. 本文来自六.维~论^文·网原文请找腾讯324,9114
Key Words: Minimum Error Thresholding；Image Thresholding；Image Segmentation；Color Image

1 引言 1
1.1 研究背景 1
1.2 本文研究的主要内容 2
2 阈值分割方法综述 3
2.1 阈值分割方法 3
2.2 阈值分割方法分类 3
3 基于灰度图像的最小误差阈值法 6
3.1 一维最小误差阈值法 6
3.1.1 基础理论 6
3.1.2 算法原理 7
3.1.3 算法流程 9
3.2 二维最小误差阈值法 9
3.2.1 算法原理 10
3.2.2 算法流程 13
3.3 二维最小误差阈值改进算法 14
3.3.1 算法原理 14
3.3.2 算法流程 16
3.4 二维直线型最小误差阈值法 17
3.4.1 算法原理 17
3.4.2 算法流程 18
4 最小误差阈值法应用于彩色图像 20
4.1 RGB彩色模型 20
4.2 一维矢量阈值分割法 21
4.3 二维矢量阈值分割法 23
5 实验结果与分析 25
5.1 阈值分割质量评价 25
5.1.1 灰度图像分割质量评价 25
5.1.2 彩色图像分割质量评价 26
5.2 阈值分割结果及分析 26
5.2.1 灰度图像分割结果及分析 26
5.2.1 彩色图像分割结果及分析 30
6 结束语 33

[1] [2] [3] [4] [5] [6] 下一页