摘 要
隨著信息技術(shù)的發(fā)展和視頻監(jiān)控系統(tǒng)的普及，視覺(jué)監(jiān)控技術(shù)在科學(xué)研究、工業(yè)生產(chǎn)中得到了越來(lái)越多的應(yīng)用。智能監(jiān)控系統(tǒng)包括三方面內(nèi)容：運(yùn)動(dòng)檢測(cè)、運(yùn)動(dòng)方向判定和圖像跟蹤，其中運(yùn)動(dòng)檢測(cè)是重要的基礎(chǔ)內(nèi)容。
論文介紹了系統(tǒng)設(shè)計(jì)中所應(yīng)用的中值濾波、灰度化、圖像增強(qiáng)、邊緣檢測(cè)等數(shù)字圖像處理、形態(tài)學(xué)等方面的相關(guān)理論及一些可行性算法。在此基礎(chǔ)上，本文針對(duì)視頻監(jiān)控系統(tǒng)中的運(yùn)動(dòng)檢測(cè)及定位進(jìn)行探討，并分三個(gè)模塊設(shè)計(jì)了一個(gè)系統(tǒng)用于檢測(cè)物體的運(yùn)動(dòng)情況和繪制運(yùn)動(dòng)軌跡。此系統(tǒng)能夠自動(dòng)提取背景，并準(zhǔn)確識(shí)別物體，處理時(shí)間短，對(duì)圖片噪聲有很強(qiáng)的抑制能力，人性化設(shè)計(jì)使界面友好，操作簡(jiǎn)單?；谝曨l的運(yùn)動(dòng)檢測(cè)研究可以應(yīng)用于不同領(lǐng)域，如交通管理中車(chē)牌的檢測(cè)、大型超市等的監(jiān)測(cè)系統(tǒng)以及刑偵方面的各種動(dòng)態(tài)識(shí)別等，除此之外，運(yùn)動(dòng)檢測(cè)系統(tǒng)的相應(yīng)功能可以作為一個(gè)監(jiān)控系統(tǒng)的軟件核心，應(yīng)用到銀行、賓館、以及珠寶行中，再者由于本系統(tǒng)可以單獨(dú)使用，所以也可以作為數(shù)字圖像處理工具使用，對(duì)電影中截取的一系列圖片進(jìn)行分析，得到想要的物體的運(yùn)動(dòng)軌跡。本系統(tǒng)采用VC2005作為開(kāi)發(fā)工具，以Windows作為圖像處理實(shí)施操作的運(yùn)行平臺(tái)，制作了數(shù)字圖像運(yùn)動(dòng)檢測(cè)系統(tǒng)。并且介紹了運(yùn)動(dòng)檢測(cè)系統(tǒng)中用到的數(shù)字圖像處理技術(shù)及詳細(xì)的步驟。伴隨著計(jì)算機(jī)技術(shù)的迅猛發(fā)展，圖像處理與我們的工作和生活方式息息相關(guān)，這門(mén)邊緣技術(shù)將得到更為廣泛的應(yīng)用。

關(guān)鍵詞：方向判定；圖像處理；識(shí)別物體；提取背景；運(yùn)動(dòng)檢測(cè)








Digital Image Processing Surveillance System

Abstract
With the development of information technology and the popularity of video monitoring system, video surveillance is widely used in science research and industry production. There are three aspects in intelligent surveillance system: motion detecting that is the essential factor, motion positioning and image tracking. And motion detection is the basic one.
Some theoretical knowledge about Digital Image Processing, Mathematical Morphology, such as median filter, gray processing, image enhancement, edge detection, et al, are discussed in this dissertation. And then, some problems about the motional detection and position in secure surveillance system are also studied. The system is designed in three modules for locating objects and drawing motion track-curve. The system has the features such as the background obtained and the object identified automatically, less demand of computation time, and noise restrained effectively. The motion detection based on video image can be used in different fields. For example, car brand detection in traffic management, surveillance system in supermarket and dynamic identify for criminal spy, etc. Besides the author also mentioned the function and the value of the sport examining system. For instance the system could use as an important part of the examine system in the software, which the bank, hotel and supermarket could make use of .Because of the sport examine system could be alone used; people could make use of the system to do with the picture. For example, the system could analysis a series of picture which grabbed from the film .By doing this, an orbit could be taken through the sport examine system .which were used the VC++ as its development instrument, the Windows as its moving platform. Meanwhile author introduced the technique which the sport examining system used. In this article author introduced the detailed sequence of the system as well. In pace with the computer technology development quickly, digital image processing will become very important in our daily lives. So I believe this technology will be made used by more people in future.

Key words：Direction Determining；Digital Image Processing；Recognition Object；Background Obtained；Motion Detection


目 錄
摘 要 I
Abstract III
1 緒論 1
1.1 數(shù)字圖像運(yùn)動(dòng)檢測(cè)國(guó)內(nèi)外發(fā)展的現(xiàn)狀 1
1.2 數(shù)字圖像處理簡(jiǎn)介 2
1.3 數(shù)字圖像處理的基本要求 3
1.4 數(shù)字圖像處理的應(yīng)用 4
2 相關(guān)技術(shù)介紹 6
2.1 Windows位圖和調(diào)色板 6
2.1.1 圖像存儲(chǔ)的調(diào)色板技術(shù) 6
2.1.2 Bmp位圖文件格式 7
2.2 圖像的平滑銳化 8
2.2.1 圖像的平滑 8
2.2.2 中值濾波 9
2.2.3 銳化 11
3 數(shù)字圖像運(yùn)動(dòng)檢測(cè)概要設(shè)計(jì) 12
3.1 性能需求 12
3.2 功能需求 12
3.3 開(kāi)發(fā)語(yǔ)言與平臺(tái) 12
3.3.1 VC2005 12
3.3.2 MFC 13
3.4 系統(tǒng)設(shè)計(jì) 17
3.4.1 界面設(shè)計(jì) 17
3.4.2 功能設(shè)計(jì) 17
4 數(shù)字圖像運(yùn)動(dòng)檢測(cè)的詳細(xì)設(shè)計(jì) 24
4.1 程序?qū)崿F(xiàn) 24
4.1.1 圖像數(shù)據(jù)讀取 24
4.1.2 圖像的中值濾波 26
4.1.3 圖像灰度化 28
4.1.4 圖像梯度的獲取 30
4.1.5 圖像特征區(qū)域的確定 32
4.1.6 得到特征區(qū)域的顏色信息 34
4.1.7 兩幅圖像的匹配 35
4.1.8 特征區(qū)域的第二次判定 40
4.1.9 物體的區(qū)域定位 40
4.1.10 圖像二值化 42
4.1.11 圖像的膨脹 43
4.1.12 圖像邊緣求取 45
4.1.13 圖像細(xì)化 46
4.1.14 圖像的去離散點(diǎn)的操作 48
4.1.15 物體中心點(diǎn)的取得 49
4.1.16 物體運(yùn)動(dòng)軌跡的繪制 50
結(jié) 論 53
參 考 文 獻(xiàn) 54
致 謝 56