【摘要】：近年來,隨著社會的發(fā)展,科學技術(shù)的進步,監(jiān)控技術(shù)的應(yīng)用在保證社會穩(wěn)定安全發(fā)展中的作用也越來越被突顯出來。在目前的相關(guān)監(jiān)控技術(shù)中,大多數(shù)傳統(tǒng)的視頻監(jiān)控系統(tǒng)仍然需要人對視頻內(nèi)容進行分析辨別,這造成了極大的資源浪費,而且隨著監(jiān)控技術(shù)被應(yīng)用的領(lǐng)域越來越廣,人們對于監(jiān)控系統(tǒng)的智能化的需求也日益迫切。 本文在深入研究了監(jiān)控系統(tǒng)的智能化相關(guān)實現(xiàn)技術(shù)以及圖像處理技術(shù)后,以智能視頻監(jiān)控系統(tǒng)在監(jiān)控區(qū)域?qū)θ四樳M行識別,以及在識別出人臉后對采集的人臉圖像進行縮放,由此解決監(jiān)控人員的對視頻內(nèi)容分析辨別困難的問題作為本文研究的重點。圍繞TI公司所開發(fā)的新型數(shù)字信號處理開發(fā)板TMS320DM6446設(shè)計了一套基于圖像識別技術(shù)的智能視頻模塊,并依據(jù)此視頻模塊的架構(gòu)構(gòu)建一套智能視頻監(jiān)控系統(tǒng)方案。 本系統(tǒng)在硬件方面以TMS320DM6446為核心,針對智能視頻監(jiān)控系統(tǒng)的需求,分別對,視頻采集模塊、數(shù)據(jù)存儲模塊、視頻后端處理模塊、視頻傳輸模塊、開關(guān)量控制模塊、以及電源供電模塊進行了相關(guān)硬件設(shè)計。使本系統(tǒng)達到性能與成本的最佳平衡。 本系統(tǒng)在軟件方面基于人臉識別原理與圖像縮放技術(shù),采用識別正確率較高的AdaBoost算法和圖像縮放效果較好的縫隙剪裁技術(shù)(Seam Carving)設(shè)計了系統(tǒng)的主要軟件框架,并分析了實現(xiàn)軟件模塊的關(guān)鍵代碼。并在通過軟件仿真驗證本系統(tǒng)所設(shè)計的相關(guān)軟件模塊能夠取得較為良好的人臉識別與圖像縮放效果。
[Abstract]:In recent years, with the development of society and the progress of science and technology, the application of monitoring technology in ensuring social stability and security development has become more and more prominent. At present, most of the traditional video surveillance systems still need people to analyze and distinguish the video content, which has caused a great waste of resources, and with the application of monitoring technology has become more and more extensive. The demand for intelligent monitoring system is becoming more and more urgent. In this paper, the intelligent correlation technology and image processing technology of the monitoring system are deeply studied, then the face is recognized by the intelligent video surveillance system in the monitoring area, and the collected face image is scaled after the recognition of the face. It is the focus of this paper to solve the problem of video content analysis and discrimination. A set of intelligent video module based on image recognition technology is designed around TMS320DM6446, a new digital signal processing board developed by TI, and a set of intelligent video surveillance system scheme is constructed according to the structure of this video module. This system takes TMS320DM6446 as the core in hardware, aiming at the demand of intelligent video surveillance system, respectively, video acquisition module, data storage module, video back-end processing module, video transmission module, switch control module, etc. And the related hardware design of power supply module. Make the system achieve the best balance between performance and cost. Based on the principle of face recognition and image scaling technology, this system designed the main software framework of the system by using the AdaBoost algorithm with high recognition accuracy and the slot cutting technology (Seam Carving), which has good image scaling effect. The key code to realize the software module is analyzed. The software simulation shows that the software module designed in this system can achieve better face recognition and image scaling effect.
【學位授予單位】：中北大學
【學位級別】：碩士
【學位授予年份】：2014
【分類號】：TN948.6;TP391.41

【參考文獻】

相關(guān)期刊論文 前4條

1 徐伯慶,孫國強,陳離;數(shù)字圖像縮放的原理與硬件實現(xiàn)[J];上海理工大學學報;2002年03期

2 王文寧;李慧娟;師磊;;一種基于顏色和形狀特征的人臉檢測方法[J];計算機系統(tǒng)應(yīng)用;2008年07期

3 白艷茹;劉賀平;朱珍民;羅海勇;先明春;高玉之;;基于TMS320DM6446的多媒體視訊終端[J];儀表技術(shù)與傳感器;2008年08期

4 王華斌;;C64+對C64內(nèi)核的性能提升[J];電子設(shè)計應(yīng)用;2006年06期

，

本文編號：2166353