【正文】 re extracted and preprocessing, get a clear face image。Finally to get the characteristic values of processing, according to the weighted Euclidean distance to calculate the similarity degree, so as to realize video mining based on human system has been done can preliminarily based on the characters of work video mining, but needs to improve its accuracy. Key words: video mining, facial feature extraction, geometric feature recognition 基于人物的相關(guān)網(wǎng)絡(luò)視頻挖掘 3 目錄 第一章 緒論 ............................................................................................................... 5 背景 ....................................................................................................................... 5 算法簡介 ................................................................................................................ 5 論文主要內(nèi)容 ........................................................................................................ 6 論文結(jié)構(gòu)安排 ........................................................................................................ 6 第二章 系統(tǒng)的概要設(shè)計(jì) ............................................................................................ 6 系統(tǒng)的解決方案 ..................................................................................................... 6 系統(tǒng)的流程圖 ........................................................................................................ 7 視頻預(yù)處理方案 ..................................................................................................... 7 人臉檢測方案 ........................................................................................................ 7 視頻挖掘方案 ........................................................................................................ 8 第三章 視頻預(yù)處理 .................................................................................................... 8 視頻預(yù)處理 ............................................................................................................ 8 幀提取 ................................................................................................................... 8 光線補(bǔ)償技術(shù) ........................................................................................................ 9 高斯平滑技術(shù) ........................................................................................................ 9 圖片灰度化 ............................................................................................................ 9 灰度均衡技術(shù) ........................................................................................................ 9 對比度增強(qiáng)技術(shù) ..................................................................................................... 9 圖片尺寸縮放 ........................................................................................................ 9 第四章 特征提取 ....................................................................................................... 9 特征提取 .............................................................................................................. 10 特征提取方案 ...................................................................................................... 10 分類器算法 ................................................................................................... 10 特征 .............................................................................................................. 10 算法 ....................................................................................................... 11 特征提取 .............................................................................................................. 12 特征處理（標(biāo)準(zhǔn)化特征值） ................................................................................ 12 第五章 視頻挖掘 ..................................................................................................... 12 視頻挖掘 .............................................................................................................. 13 基于人物的相關(guān)網(wǎng)絡(luò)視頻挖掘 4 視頻挖掘方案 ...................................................................................................... 13 標(biāo)準(zhǔn)化特征值計(jì)算 ............................................................................................... 13 數(shù)據(jù)庫檢索 .......................................................................................................... 14 獲取對應(yīng)視頻文件 ............................................................................................... 14 第六章 系統(tǒng)測試及運(yùn)行結(jié)果 ................................................................................... 14 測試原則 ....................................................................