【正文】 進(jìn)行分析說(shuō)明 。 論文內(nèi)容及組織安排 圖像處理技術(shù)領(lǐng)域突飛猛進(jìn)的發(fā)展，人們對(duì)能夠提供更多信息的視頻圖像越來(lái)越該興趣，對(duì)視頻圖像或圖像序列處理的核心技術(shù) — 圖像檢測(cè)技術(shù)的快速實(shí) 用性、準(zhǔn)確性、 穩(wěn)定 性等的要求也在不斷的提高，論文對(duì)目前廣泛采用的實(shí)時(shí)圖像檢測(cè)方法的性能進(jìn)行了理論分析和 比較 ，并 針對(duì) 簡(jiǎn)單背景視頻 和復(fù)雜背景 視頻 利用背景差分法和改進(jìn)的背景差分法進(jìn)行實(shí)時(shí)檢測(cè) ， 得到 較好的實(shí)驗(yàn)效果。有的雖然 準(zhǔn)確性 較高，但算法過(guò)于復(fù)雜，不利于進(jìn)行實(shí)時(shí)應(yīng)用。 在現(xiàn)實(shí)生活中，大量有意義的視覺(jué)信息都包含在運(yùn)動(dòng)當(dāng)中，盡管人們可以通過(guò)眼睛感知運(yùn)動(dòng)物體的存在，但是對(duì)于機(jī)器來(lái)說(shuō)，從成像設(shè)備獲取物體的運(yùn)動(dòng)信息則是一個(gè)難題。 國(guó)內(nèi)方 面，近年來(lái)中國(guó)圖像圖形學(xué)學(xué)會(huì)、 中國(guó)科學(xué)院等對(duì)運(yùn)動(dòng)方面的研究給予 很大關(guān)注，全國(guó)許多大學(xué)和科研結(jié)構(gòu)也都競(jìng)相展開對(duì) 運(yùn)動(dòng) 目標(biāo)檢測(cè)的研究 [6]。 80 年代運(yùn)用二維影像光流法、梯 度法能夠處理小運(yùn)動(dòng)范圍的移動(dòng)目標(biāo)檢測(cè)。 現(xiàn)在 運(yùn)動(dòng)目標(biāo)檢測(cè) 技術(shù) 己經(jīng)被廣泛地應(yīng)用到多媒體技術(shù)、視頻數(shù)據(jù)壓縮技術(shù)、工業(yè)生產(chǎn)、軍事應(yīng)用等各個(gè)領(lǐng)域 [4]。 運(yùn)動(dòng)目標(biāo)檢測(cè)是計(jì)算機(jī)視覺(jué)研究領(lǐng)域中的一個(gè)重要且困難的課題，在實(shí)現(xiàn)運(yùn)動(dòng)補(bǔ)償、視頻壓縮編碼、視頻理解時(shí)都需要用到運(yùn)動(dòng)目標(biāo)檢測(cè)技術(shù)。圖像處理從靜止圖像轉(zhuǎn)移到圖像序列上，通過(guò)對(duì)多幀圖像分析，可獲得從單一圖像中不可能得到的信息 ， 只有在圖像序列上我們才有可能認(rèn)識(shí)和分析動(dòng)態(tài)過(guò)程 [2]。 關(guān)鍵詞： 視頻監(jiān)控、 運(yùn)動(dòng)目標(biāo)檢測(cè)、背景差分法、 背景提取、 特征匹配 Abstract Motion target detection is one of the main issues in puter vision, which bines advanced technologies in image processing, Pattern recognition, automatic control, artificial intelligence, puter and other relative fields. It has broadly applied in military visual missile guidance, video surveillance, medical image analysis, intelligent transportation and other fields, so this project research has important theoretical significance and practical value. Based on the summary and analysis of the current detecting moving objects methods, this paper has designed and implemented an experiment which can detect moving object in image sequence after captured, using the detecting algorithm of realtime for video, which obtain a better experimental result. In the collection of images, the existed AVI video taken by myself has been adopted in this paper and transformed properly. In that way, it can facilitate the reading and transformation of MATLAB. In the process of object detection, this paper also offers simple background and plex background for object detection. In terms of these two AVI videos, detection to the AVI video via utilizing background subtraction and the improved background subtraction, to some extent, can enhance the parability of the experimental Characteristics including object positions and so on, given by the operation of this experiment have enhanced the intuitiveness of the experiment. The experimental result shows that the detecting algorithm used in this system attain the expectable purpose. The algorithm and methods used in this paper will be valued either in the domain of intelligent video control, air surveillance, and industrial detection and so on. Keywords: intelligent monitoring, moving object detection, background subtraction, background extraction, feature matching i 目 錄 摘要 Abstract 第一章 緒論 .................................................................................................................. 1 課題研究背景及意義 ..................................................................................... 1 國(guó)內(nèi)外研究現(xiàn)狀 ............................................................................................. 2 論文內(nèi)容及組織安排 ..................................................................................... 3 第二章 運(yùn)動(dòng)目標(biāo)檢測(cè)有關(guān)的理論基礎(chǔ) ...................................................................... 4 數(shù)字圖像基本 概念 ......................................................................................... 4 圖像灰度變換 ................................................................................................. 5 中值濾波 ......................................................................................................... 6 邊緣檢測(cè) ......................................................................................................... 6 圖像配準(zhǔn) ......................................................................................................... 7 運(yùn)動(dòng)目標(biāo)檢測(cè)技術(shù) ......................................................................................... 7 幀間差分法 ........................................................................................... 7 背景差分法 ........................................................................................... 8 光流法 ................................................................................................... 8 統(tǒng)計(jì)學(xué)法 ............................................................................................... 8 第三章 運(yùn)動(dòng)目標(biāo)檢測(cè)設(shè)計(jì)分 析 ................................................................................ 10 總體設(shè)計(jì)思路 ............................................................................................... 10 背景差分法 ................................................................................................... 10 總體設(shè)計(jì)流程圖 ................................................................................ 10 功能模塊的分析 ................................................................................ 11 改進(jìn)的背景差分法 ....................................................................................... 12 總體設(shè)計(jì)流程圖 ................................................................................ 12 功能模塊的分析 ................................................................................ 12 第四章 運(yùn)動(dòng)目標(biāo)檢測(cè)實(shí)驗(yàn)驗(yàn)證與結(jié)果分析 ............................................................ 15 運(yùn)動(dòng)目標(biāo)檢測(cè)實(shí)驗(yàn)介紹 ............................................................................... 15 簡(jiǎn)單背景視頻 ............................................................................................... 15 ii 背景差分法實(shí)現(xiàn) ................................................................................ 15 改進(jìn)的背景差分法實(shí)現(xiàn) .................................................................... 17 復(fù)雜背景視頻 ............................................................................................... 18 背景差分法實(shí)現(xiàn) ................................................................................ 18 改進(jìn)的背景差分法實(shí)現(xiàn) .................................................................... 20 結(jié)果分析 ....................................................................................................... 21 第五章 總結(jié)與展望 .................................................