【正文】 畢業(yè)論文 基于塊匹配的人群運(yùn)動(dòng)估計(jì) 基于塊匹配的人群運(yùn)動(dòng)估計(jì) 軟件工程 學(xué)生 舒禹銘 指導(dǎo)老師 李曉華 [摘要 ] 智能化人群監(jiān)控是智能視頻監(jiān)控研究中的一個(gè)重要課題，它作為智能監(jiān)控中的一項(xiàng)關(guān)鍵技術(shù)，在人群管理、公共場(chǎng)所設(shè)計(jì)、虛擬環(huán)境建模、視覺(jué)監(jiān)控、智能環(huán)境模擬等方面都有著重要的應(yīng)用價(jià)值。 隨著經(jīng)濟(jì)社會(huì)的發(fā)展，各種公共場(chǎng)地和設(shè)施中的人群流動(dòng) 越來(lái)越頻繁。智能化人群監(jiān)控技術(shù)應(yīng)運(yùn)而生，它主要包括人群的密度估計(jì)和運(yùn)動(dòng)估計(jì)兩部分內(nèi)容。它們能夠直接或間接地提高上述場(chǎng)合工作人員的工作效率和建筑設(shè)施的利用率，因此對(duì)人群密度估計(jì)和運(yùn)動(dòng)估計(jì)方法的研究有著深遠(yuǎn)的意義和廣闊的前景。 [主題詞 ] 人群監(jiān)控；人群運(yùn)動(dòng)估計(jì)；塊匹配； OpenCV 四川大學(xué)本科畢業(yè)論文 基于塊匹配的人群運(yùn)動(dòng)估計(jì) 3 Crowd motion estimation based on BMA Software Engineering Student： YuMingshu Adviser: XiaoHuali [Abstract] Intelligentized crowd surveillance tecllIlology is an important research sub—field in intelligem video surveillance system． As a key tecllnology of intelligent surveillarlce， it is of great value in a large number of applications such as crowd management，public space design virtual environments simulate， visual surveillanlce， intelligent enviroments aIlalysis,etc． With the development of economic society, the crowd in various kinds of public places flows more and more frequently. How to manage and control the crowd effectively es to be an important issue which we have to consider nowadays. Intelligentized crowd surveillance technology arises at the very moment. It mainly includes both density estimation and motion paper set about the crowd motion estimation. Intelligentized crowd motion estimation can be used for monitoring and managing the crowd, at the same time, it can also be used for market survrey in the mercial field， traffic safety and architectural design field,etc. it can help staff members in the above mentioned occasions improve working efficiency and improve utilization ratio of building facilities directly or indirectly,so there is farreaching meaning and wide prospect in crowd’s density and motion estimation research. According to OpenCV, this paper use BMA(Block Matching Algorithm) for crowd motion estimation， and before the function running ,it has a specific analysis about OpenCV and the important case of BMA.. [Key Words] crowd surveillance； crowd motion estimation； Block Matching Algorithm；OpenCV 四川大學(xué)本科畢業(yè)論文 基于塊匹配的人群運(yùn)動(dòng)估計(jì) 目 錄 1 緒論 ............................................... 1 研究背景 ................................................ 1 智能監(jiān)控 ......................................................... 1 人群監(jiān)控的提出 ................................................... 1 運(yùn)動(dòng)估計(jì) ......................................................... 2 塊匹配算法 ....................................................... 2 OpenCV ........................................................... 2 論文工作構(gòu)思 ..................................................... 3 國(guó)內(nèi)外研究與技術(shù)現(xiàn)狀 ..................................... 3 智能人群監(jiān)控的研究現(xiàn)狀 ........................................... 3 運(yùn)動(dòng)估計(jì)方法的研究現(xiàn)狀 ........................................... 4 塊匹配現(xiàn)狀 ....................................................... 4 論文主要工作 ............................................ 5 論文組織與結(jié)構(gòu) ........................................... 5 2 塊匹配算法介紹及分析 ................................. 6 運(yùn)動(dòng)估計(jì) ................................................ 6 塊匹配基本思想 ........................................... 6 初始搜索點(diǎn)的選擇 ................................................. 7 塊匹配準(zhǔn)則 ....................................................... 8 搜索策略 ......................................................... 8 典型的塊匹配算法 ......................................... 9 各模塊擬采用的算法 ...................................... 15 3 人群運(yùn)動(dòng)估計(jì)的塊匹配算法實(shí)現(xiàn) ......................... 16 實(shí)現(xiàn)工具 OPENCV ........................................ 16 ............................................... 17 CvPoint,CvSize .................................................. 17 CvMat ........................................................... 17 IplImage ........................................................ 19 其他函數(shù) ........................................................ 21 .................................................... 24 cvCvtColor ...................................................... 24 cvSmooth ........................................................ 26 cvCalcOpticalFlowBM ............................................. 27 ............................................... 28 四川大學(xué)本科畢業(yè)論文 基于塊匹配的人群運(yùn) 動(dòng)估計(jì) 2 CvCapture ....................................................... 28 窗口函數(shù) ........................................................ 28 cvWaitKey ....................................................... 31 算法實(shí)現(xiàn) ............................................... 31 圖像處理（視頻處理） ............................................ 32 塊匹配 .......................................................... 37 過(guò)濾非運(yùn)動(dòng)的物體 ................................................ 39 連線 ............................................................ 40 程序算法流程圖 .......................................... 41 程序?qū)崿F(xiàn)截圖 ........................................... 42 4 軟件的測(cè)試 ......................................... 45 測(cè)試環(huán)境 ............................................... 45 硬件環(huán)境 ........................................................ 45 軟件環(huán)境 ........................................................ 45 測(cè)試步驟 ............................................... 45 測(cè)試總體目標(biāo) .................................................... 45 測(cè)試計(jì)劃實(shí)施步驟 ................................................ 45 測(cè)試用例 ............................................... 45 分支測(cè)試 ........................................................ 45 集成測(cè)試 ........................................................ 55 測(cè)試結(jié)果分析 ........................................... 57 5 總結(jié)和展望 ......................................... 58 工作總結(jié) ............................................... 58 心得體會(huì) ............................................... 58 致 謝 .................................