【摘要】：視頻監(jiān)控是一種通過數(shù)據(jù)傳輸鏈路將采集的視頻、音頻等數(shù)據(jù)信息傳輸給信號處理服務(wù)器,并經(jīng)由服務(wù)器處理發(fā)送到對應(yīng)的輸出終端的技術(shù)。視頻監(jiān)控可以進(jìn)行實時性監(jiān)測,并對監(jiān)測結(jié)果進(jìn)行存儲,方便事后調(diào)查取證。視頻監(jiān)控技術(shù)的發(fā)展要求隨著信息化的發(fā)展越來越高,視頻監(jiān)控不僅需要滿足可控可管,還需要較高的魯棒性,方便監(jiān)控系統(tǒng)升級改造。當(dāng)前技偵視頻監(jiān)控系統(tǒng)設(shè)計單一,可管性較低,并且監(jiān)控視頻利用率較低,不能及時進(jìn)行有效檢索和目標(biāo)追蹤。本文針對視頻監(jiān)控系統(tǒng)設(shè)計單一,視頻檢索效率較低的問題,進(jìn)行了探討和研究。主要工作如下:1.根據(jù)網(wǎng)絡(luò)視頻監(jiān)控系統(tǒng)的優(yōu)勢,結(jié)合多層級網(wǎng)絡(luò)模型,通過視頻監(jiān)控點映射目錄樹的方法對視頻監(jiān)控系統(tǒng)實現(xiàn)多層多級管理體系,并設(shè)計了對應(yīng)該體系的目錄標(biāo)識碼。2.在視頻檢索系統(tǒng)設(shè)計中,首先根據(jù)視頻幀的背景差分法確定視頻幀的背景畫面,然后對檢測視頻進(jìn)行直方圖距離差計算以確定視頻畫面中變化的部分,最后根據(jù)直方圖距離差的大小確定目標(biāo)檢索結(jié)果。3.設(shè)計了一種目標(biāo)追蹤的方法。該方法根據(jù)目標(biāo)的運(yùn)動速度和方向進(jìn)行預(yù)測,并判斷出目標(biāo)出現(xiàn)的區(qū)域,方便視頻監(jiān)控前端設(shè)備進(jìn)行準(zhǔn)確及時的切換,從而實現(xiàn)無縫銜接。4.文中視頻監(jiān)控系統(tǒng)工程實現(xiàn)中,主要實現(xiàn)工作集中在對視頻圖像采集和視頻傳輸上,并在此基礎(chǔ)上對視頻檢索系統(tǒng)進(jìn)行了測試。攝像頭驅(qū)動的設(shè)計主要采用Linux操作系統(tǒng)自帶的驅(qū)動程序,文中對配置進(jìn)行修改,然后在內(nèi)核中編譯驅(qū)動;圖像采集程序的設(shè)計采用Video Linux設(shè)計實現(xiàn);視頻數(shù)據(jù)傳輸設(shè)計使用基于IPv4的UDP協(xié)議。測試結(jié)果表明,文中所設(shè)計的視頻監(jiān)控系統(tǒng)性能達(dá)到要求,所提出的視頻檢索方法檢索成功率高,并能有效追蹤目標(biāo),具有較高的有效性。視頻監(jiān)控系統(tǒng)相當(dāng)于公安機(jī)關(guān)全天候的眼睛,已經(jīng)成為保障社會安全、維護(hù)公共利益、促進(jìn)社會穩(wěn)定、打擊違法犯罪、管理社會治安的必備武器,有效地利用視頻監(jiān)控系統(tǒng)可以較大幅度提高社會穩(wěn)定。
[Abstract]:Video surveillance is a kind of technology that transmits the collected video, audio and other data information to the signal processing server through the data transmission link and sends to the corresponding output terminal through the server processing. Video surveillance can be used for real-time monitoring, and the monitoring results can be stored to facilitate the investigation and collection of evidence after the event. With the development of information technology, video surveillance needs not only controllable control, but also high robustness, which is convenient to upgrade and transform the monitoring system. At present, the video surveillance system of technical reconnaissance is single, low manageability and low utilization ratio of surveillance video, so it can not carry out effective retrieval and target tracking in time. In this paper, the problem of single design of video surveillance system and low efficiency of video retrieval is discussed and studied. The main work is as follows: 1. According to the advantages of the network video surveillance system and the multi-level network model, the multi-layer and multi-level management system of the video surveillance system is implemented by the method of video monitoring point mapping directory tree, and the directory identification code of the corresponding system is designed. 2. In the design of the video retrieval system, the background picture of the video frame is first determined according to the background differential method of the video frame, and then the histogram distance difference is calculated to determine the changing part of the video picture. Finally, the target retrieval results are determined according to the distance difference of histogram. 3. 3. A method of target tracking is designed. The method predicts the moving speed and direction of the target, and determines the region of the target, which facilitates the video surveillance front-end equipment to switch accurately and timely, thus realizing seamless connection. 4. In the project implementation of video surveillance system, the main work is focused on video image acquisition and video transmission, and on the basis of this, the video retrieval system is tested. The design of camera driver is mainly based on the driver of Linux operating system. The configuration is modified, then the driver is compiled in the kernel, and the design of image acquisition program is implemented by Video Linux. The design of video data transmission uses UDP protocol based on IPv4. The test results show that the performance of the video surveillance system designed in this paper meets the requirements, the proposed video retrieval method has a high success rate, and can effectively track the target, which is highly effective. The video surveillance system is equivalent to the all-weather eyes of the public security organs. It has become an essential weapon to safeguard social security, safeguard the public interest, promote social stability, crack down on illegal crimes, and manage public order. Effective use of video surveillance system can greatly improve social stability.
【學(xué)位授予單位】：浙江工業(yè)大學(xué)
【學(xué)位級別】：碩士
【學(xué)位授予年份】：2015
【分類號】：TN948.6

【參考文獻(xiàn)】

相關(guān)期刊論文 前10條

1 齊興斌;趙麗;李雪梅;田濤;;視頻檢索中基于GMM聚類的無監(jiān)督情感場景檢測[J];電視技術(shù);2015年05期

2 鄭海波;韓小萱;史云靜;李潔;朱秀昌;;基于內(nèi)容的海量監(jiān)控視頻的多層次檢索系統(tǒng)[J];電視技術(shù);2014年19期

3 朱麗英;梁辰;張鴻洲;;基于運(yùn)動目標(biāo)的監(jiān)控視頻檢索方法[J];計算機(jī)應(yīng)用與軟件;2011年12期

4 宗文杰;余青松;張敏;黃俊;;基于GIS的視頻監(jiān)控系統(tǒng)研究[J];計算機(jī)工程與設(shè)計;2011年02期

5 黃丙湖;韓李濤;孫根云;陳龍;;三維GIS與視頻監(jiān)控系統(tǒng)的集成與應(yīng)用研究[J];測繪通報;2011年01期

6 李澤洲;歐陽建權(quán);張敏;劉煒;;基于視頻指紋的視頻片段檢索方法[J];計算機(jī)工程;2010年07期

7 李旭東;;數(shù)字化安防系統(tǒng)的應(yīng)用研究與設(shè)計[J];安防科技;2008年03期

8 許f3;王德興;劉賢慶;;視頻定位技術(shù)在城市公共安全視頻監(jiān)控系統(tǒng)中的應(yīng)用[J];梧州學(xué)院學(xué)報;2007年06期

9 陳衛(wèi)國;戴瑜興;;基于CAN總線的視頻切換矩陣設(shè)計[J];低壓電器;2007年22期

10 盧秋波;;視頻監(jiān)控技術(shù)簡介與發(fā)展趨勢[J];電信網(wǎng)技術(shù);2007年01期

相關(guān)碩士學(xué)位論文 前2條

1 王晨;智能視頻監(jiān)控系統(tǒng)設(shè)計[D];南京理工大學(xué);2007年

2 陳磊;視頻圖像中的運(yùn)動目標(biāo)檢測算法研究[D];湖南大學(xué);2007年

，

本文編號：2327960