【正文】 characteristics using SVMbased clustering algorithm. After extracting samples? characteristic value, put into SVM?s training work to have training. Finally the input image can be classified by the classification results. This paper concludes the approach to the classification of nonsupervised classification accuracy of the results meanwhile analyzes and discusses the accuracy. This paper used Kmenas classification method in the field of unsupervised classification. After preconfiguring data, put into classified image, and then by calling the classification function, the system will output the results of automatic classification. Based on algorithm, design and implementation of fruit intelligent image analysis application system with a characteristic value database, training work, image supervised classification and image unsupervised classification features. When the image has supervised classification, the SVM classification count classification, the accuracy rate can reach nearly 70%. KEY WORDS SVM kmeans Image classification I 目 錄 第一章 緒論 .............................................................................................................................. 1 智能圖像分析概述 .................................................................................................... 1 課題背景 ............................................................................................................ 1 國(guó)內(nèi)外研究現(xiàn)狀 ................................................................................................ 2 聚類(lèi)分析 .................................................................................................................... 3 課題目標(biāo)及本文研究?jī)?nèi)容 ........................................................................................ 3 預(yù)期目標(biāo) ............................................................................................................ 3 主要研究?jī)?nèi)容 .................................................................................................... 3 系統(tǒng)方案 ............................................................................................................ 4 本文的結(jié)構(gòu) ........................................................................................................ 4 第二章 技術(shù)基礎(chǔ) ...................................................................................................................... 5 圖像特征 .................................................................................................................... 5 圖像分類(lèi)方法 ............................................................................................................ 5 圖像分類(lèi)概念 .................................................................................................... 5 圖像分類(lèi)原理 .................................................................................................... 6 圖像分類(lèi)方法 .................................................................................................... 6 MatLab 及圖像智能處理工具箱 .............................................................................. 7 第三章 圖像矩不變特征提取 .................................................................................................. 9 圖像矩不變特征介紹 .............................................................................................. 11 圖像矩不變特征提取 .............................................................................................. 12 第四章 分類(lèi)算法 .................................................................................................................... 14 SVM 分類(lèi)算法 ........................................................................................................ 14 kmeans 分類(lèi)算法 ................................................................................................... 16 第五章 基于 MatLab的圖像分析軟件實(shí)現(xiàn) ........................................................................ 19 軟件功能及系統(tǒng)流程 .............................................................................................. 19 關(guān)鍵 函數(shù)詳述 .......................................................................................................... 19 II 圖像灰度化 ...................................................................................................... 19 圖像平滑與圖像銳化 ...................................................................................... 20 中值濾波 .......................................................................................................... 20 圖像銳化 .......................................................................................................... 21 Hu 矩不變特征值 ................................................................................................ 21 SVM 神經(jīng)網(wǎng)絡(luò)的建立和訓(xùn)練 ............................................................................ 22 kmeans 分類(lèi)函數(shù) ............................................................................................... 24 第六章 系統(tǒng)測(cè)試 .................................................................................................................... 26 系統(tǒng)界面 .................................................................................................................. 26 功能測(cè)試及統(tǒng)計(jì) ...................................................................................................... 30 訓(xùn)練樣本 .......................................................................................................... 30 結(jié)果與分析 ...................................................................................................... 30 第七章 結(jié)論與展望 ................................................................................................................ 33 結(jié)果與結(jié)論 .............................................................................................................. 33 問(wèn)題與展望 .............................................................................................................. 33 心得體會(huì) ...............