本文選題：�；穫}庫 + 智能監(jiān)控　； 參考：《北方民族大學(xué)》2016年碩士論文

【摘要】：近年來,石油化工產(chǎn)業(yè)發(fā)展迅速,�；穂1]種類不斷增多,存放這些�；返膫}庫因此也越來越密集,甚至有些已經(jīng)深入一些居民區(qū),它們成為了威脅附近居民生命財產(chǎn)安全的定時炸彈。因此,如何對�；穫}庫進行管理和監(jiān)控,避免倉庫起火爆炸等事故的發(fā)生,成為安全部門和相關(guān)企業(yè)亟待解決的問題。據(jù)統(tǒng)計,人為因素依然是事故發(fā)生的主要因素,因此加強對進出�；穫}庫人員的監(jiān)控是解決問題的有效途徑。本課題利用計算機視覺技術(shù)、嵌入式技術(shù)以及傳感器技術(shù)等,在對國內(nèi)外現(xiàn)有相關(guān)系統(tǒng)進行深入分析的基礎(chǔ)上,研究與設(shè)計了一套基于ARM的�；穫}庫出入人員智能監(jiān)控軟件系統(tǒng)。本文首先介紹了智能監(jiān)控系統(tǒng)的相關(guān)理論及技術(shù),包括運動目標檢測、人臉識別以及目標跟蹤等算法,并提出了基于混合高斯背景建模與CamShift算法結(jié)合的運動目標跟蹤算法;接著,綜合分析了本系統(tǒng)的功能需求,并結(jié)合實際情況,對系統(tǒng)進行了總體分析與設(shè)計,將系統(tǒng)劃分為前端子系統(tǒng)和后端子系統(tǒng)兩大部分,前端子系統(tǒng)采用了以S3C6410 CPU芯片為核心的硬件平臺,以嵌入式Linux系統(tǒng)為核心的軟件平臺及Qt Creator等開發(fā)工具完成對監(jiān)控數(shù)據(jù)的實時采集、處理、發(fā)送等,后端則對前端子系統(tǒng)發(fā)送的數(shù)據(jù)進行顯示、處理并提供及時的預(yù)報警功能;然后,本文對系統(tǒng)的各個功能模塊進行了詳細的設(shè)計,并且介紹了軟硬件開發(fā)環(huán)境的配置與搭建,本文視頻分析模塊采用了OpenCV2編程,OpenCV是一個開源的跨平臺的計算機視覺庫,利用該視覺庫極大簡化了視頻分析模塊的開發(fā)難度;最后,本文對系統(tǒng)實現(xiàn)方式給出了簡要的介紹,并對系統(tǒng)進行了測試,測試結(jié)果表明初步實現(xiàn)了系統(tǒng)的需求。
[Abstract]:In recent years, the rapid development of petroleum and chemical industry, hazardous chemicals, [1] growing, storage of these hazardous chemicals warehouse and therefore more and more intensive, and even some have been deep in some neighborhoods, they become a bomb threat to nearby residents life and property safety. Therefore, how the hazardous goods warehouse management and monitoring. Avoid warehouse fire explosion and other accidents, become the urgent security departments and related enterprises. According to statistics, human factors are the main factors of the accident, therefore to strengthen monitoring of the import of hazardous chemicals warehouse personnel is an effective way to solve the problem. The subject of the use of computer vision technology, embedded technology and sensor technology and in-depth analysis based on the existing domestic and foreign system, research and design of a ARM based warehouse staff access intelligent monitoring software System. This paper firstly introduces the related theory and technology of the intelligent monitoring system, including the moving target detection, face recognition and target tracking algorithm, and put forward the moving target tracking algorithm combined with the mixed Gauss background modeling and algorithm based on CamShift; then, the comprehensive analysis of the functional requirements of the system, combined with the actual situation of system analysis and overall design, the system is divided into front-end system and terminal system of two parts, the front terminal system uses S3C6410 CPU chip as the core of hardware platform, the embedded Linux system as the core of the software platform and Qt Creator development tools to monitor real-time acquisition, data processing, transmission etc. the back-end of sending terminal, system data display, provide timely warning and treatment; then, the system can work on the various modules of the The detailed design, and introduces the development environment of software and hardware configuration and set up in this paper, the video analysis module uses the OpenCV2 programming, OpenCV is a cross platform computer vision library, an open source, which greatly simplifies the difficulty of the development of video analysis module using the visual library; most after the system implementation gives a brief is introduced, and the system is tested, test results show that the initial implementation of the system requirements.

【學(xué)位授予單位】：北方民族大學(xué)
【學(xué)位級別】：碩士
【學(xué)位授予年份】：2016
【分類號】：TP277
，

本文編號：1742453