基于信息熵最大的圖像分割研究

摘要
圖像分割是根據(jù)圖像的某些特征或特征集合的相似性準(zhǔn)則，對圖像像素進(jìn)行分組聚類，把圖像平面劃分為一系列“有意義”的區(qū)域，使其后的圖像分析及識(shí)別等高級處理階段所要處理的數(shù)據(jù)量大大減少，同時(shí)又保留有關(guān)圖像結(jié)構(gòu)特征的信息。在信息理論中，熵用于度量信息傳輸?shù)挠行浴?br> 本論文主要討論針對數(shù)字圖像利用信息熵標(biāo)準(zhǔn)結(jié)合閾值迭代法進(jìn)行圖像分割，以提高圖像分割的準(zhǔn)確性。設(shè)計(jì)信息熵算法并程序?qū)崿F(xiàn)，仿真結(jié)果表明編寫程序有效，分割效果明顯。
關(guān)鍵詞：閾值分割 迭代法 信息熵


Maximum entropy-based image segmentation study
Abstract

Based on the similarity criterion of the image’ some characteristics or characteristics set, image segmentation technique divides the image pixels into groups and partition the image into a series of "meaningful" regions. This technique will greatly reduce the amount of data in the following advanced processing to the image analysis and identification, as will as retain the information about image structure characteristics. In information theory, entropy is used to measure the effectiveness of information transmission.
In order to improve the accuracy of image segmentation ，this thesis makes a research on the image segmentation technique which is the combination of standard information entropy and Iteration threshold.Meanwhile this thesis designs entropy algorithm and gives its program. Simulation results show that program runs well, segmentation result is effective.
K eyw ords：Threshold segmentation Iterative Method Information entropy


目錄
摘要 I
Abstract II
1.緒論 1
1.1圖像 1
1.2圖像處理技術(shù) 2
1.3發(fā)展前景與應(yīng)用 6
2圖像分割 9
2.1定義 9
2.2分割方法簡介 10
3熵、信息熵 13
3.1熵 13
3.1.1基本特性 13
3.1.2應(yīng)用 13
3.2信息熵 15
4.基于最大信息熵算法實(shí)現(xiàn) 17
4.1 matlab簡介 17
4.2基于最大信息熵算法 20
4.2.1算法程序描述 20
4.2.1程序流程圖 21
4.2.2算法程序 22
4.2.3仿真及結(jié)果分析 26
結(jié)論 37
致謝 38
英文資料 39
6.1英文原文 39
6.2中文翻譯 46
參考文獻(xiàn) 50