發(fā)布時(shí)間：2018-02-03 23:08

  本文關(guān)鍵詞： 智慧幼兒園 視頻監(jiān)控 異常檢測(cè) 目標(biāo)檢測(cè) 目標(biāo)跟蹤 運(yùn)動(dòng)模型　出處：《浙江大學(xué)》2017年碩士論文　論文類型：學(xué)位論文

【摘要】：基于視頻監(jiān)控的智慧幼兒園安全檢測(cè),是指以視頻監(jiān)控為載體對(duì)智慧幼兒園內(nèi)幼兒的異常行為進(jìn)行實(shí)時(shí)自動(dòng)的檢測(cè),從而最大程度確保幼兒的安全。一方面,由于當(dāng)今社會(huì)正處于信息化時(shí)代,物聯(lián)網(wǎng)的存在使得我們可以獲取到海量的數(shù)據(jù);另一方面,隨著人們生活水平的提高,家長越來越關(guān)注幼兒的成長和身心健康。相比于傳統(tǒng)的幼兒園,智慧幼兒園的安全檢測(cè)也需要更加完善,因此,本文致力于智慧幼兒園內(nèi)的幼兒安全檢測(cè)關(guān)鍵技術(shù)研究,以視頻監(jiān)控為載體,實(shí)現(xiàn)對(duì)智慧幼兒園的安全檢測(cè)。按照視頻監(jiān)控場(chǎng)景的復(fù)雜度和異常檢測(cè)的要求,園內(nèi)幼兒安全檢測(cè)可分為簡單場(chǎng)景下的檢測(cè)和復(fù)雜場(chǎng)景下的檢測(cè)。首先,本文提出了越界檢測(cè)、基于封閉區(qū)域內(nèi)的規(guī)則檢測(cè)和摔倒檢測(cè)這三種基于規(guī)則的幼兒個(gè)體異常行為檢測(cè)算法,主要用于園內(nèi)簡單場(chǎng)景下幼兒的異常行為檢測(cè)。這三種檢測(cè)算法實(shí)現(xiàn)簡單,能夠取得很好的檢測(cè)效果,檢出率分別為100%、100%和92%,但需要事先人為規(guī)定異常規(guī)則,可擴(kuò)展性不強(qiáng)。其次,本文提出了一種基于改進(jìn)的Camshift/Kalman個(gè)體運(yùn)動(dòng)多目標(biāo)跟蹤算法,并根據(jù)此跟蹤算法進(jìn)一步提出了一種基于運(yùn)動(dòng)模型的幼兒個(gè)體運(yùn)動(dòng)異常檢測(cè)算法。其中,提出的個(gè)體運(yùn)動(dòng)多目標(biāo)跟蹤算法用于實(shí)現(xiàn)多目標(biāo)跟蹤、目標(biāo)跟蹤自動(dòng)初始化、目標(biāo)跟蹤自動(dòng)失效處理等功能。提出的幼兒個(gè)體運(yùn)動(dòng)異常檢測(cè)算法在利用個(gè)體運(yùn)動(dòng)多目標(biāo)跟蹤算法獲取到運(yùn)動(dòng)目標(biāo)信息后,以視頻監(jiān)控圖像中最小子塊為單位,通過建立的正常運(yùn)動(dòng)模型對(duì)未知運(yùn)動(dòng)目標(biāo)進(jìn)行異常檢測(cè),實(shí)現(xiàn)對(duì)幼兒在視頻監(jiān)控中的位置、運(yùn)動(dòng)方向、運(yùn)動(dòng)速率和停留時(shí)間等信息的實(shí)時(shí)檢測(cè)。該算法具有較高的準(zhǔn)確性和較快的檢測(cè)速度,特別是針對(duì)運(yùn)動(dòng)目標(biāo)較少的場(chǎng)景中的個(gè)體運(yùn)動(dòng)異常檢測(cè),平均召回率為92.6%,平均檢測(cè)速度為33.8 ms/幀,優(yōu)于同類個(gè)體異常檢測(cè)算法的實(shí)驗(yàn)結(jié)果數(shù)據(jù)。第三,本文提出了一種結(jié)合背景差分和光流法的幼兒群體運(yùn)動(dòng)狀態(tài)突變事件檢測(cè)算法,用于園內(nèi)幼兒集體活動(dòng)時(shí)群體運(yùn)動(dòng)狀態(tài)突變的異常檢測(cè)。該算法結(jié)合了背景差分算法數(shù)據(jù)處理速度快的優(yōu)勢(shì)和光流算法實(shí)驗(yàn)結(jié)果準(zhǔn)確性高的優(yōu)勢(shì),對(duì)于UMN視頻數(shù)據(jù)庫中的視頻片段AUC值可以達(dá)到0.96以上,平均檢測(cè)速度為25.8 ms/幀,相比于其它同類群體異常檢測(cè)算法在綜合性能上有明顯的提高。最后,利用基于像素統(tǒng)計(jì)的人群密度估算方法,將以上提出的各種安全檢測(cè)算法整合在一起,組合成一套比較完整的適用于智慧幼兒園內(nèi)的視頻監(jiān)控安全檢測(cè)算法,并利用OpenCV開源計(jì)算機(jī)視覺庫,在 Windows 操作系統(tǒng) Microsoft Visual Studio 2010 開發(fā)平臺(tái)下用 Visual C++語言對(duì)智慧幼兒園視頻監(jiān)控安全檢測(cè)關(guān)鍵技術(shù)進(jìn)行了實(shí)現(xiàn)。
[Abstract]:Intelligent Kindergarten Security Detection based on Video Surveillance refers to the use of video surveillance as a carrier to carry out real-time and automatic detection of abnormal behavior of children in intelligent kindergartens so as to ensure the safety of children to the greatest extent. On the one hand. As the society is now in the information age, the existence of the Internet of things enables us to obtain a large amount of data; On the other hand, with the improvement of people's living standards, parents are paying more and more attention to the growth and physical and mental health of their children. This paper is devoted to the research of key technology of infant safety detection in intelligent kindergarten, and realizes the security detection of intelligent kindergarten with video surveillance as the carrier, according to the complexity of video surveillance scene and the requirements of abnormal detection. The safety detection of children in the garden can be divided into simple scene detection and complex scene detection. Firstly, this paper proposes the cross boundary detection. Rule detection and fall detection are three rule-based algorithms for individual abnormal behavior detection in closed areas. Mainly used for the detection of abnormal behavior of children in the simple scene of the garden. These three detection algorithms are simple and can achieve a good detection effect. The detection rate is 100% and 92% respectively. However, the outlier rules need to be stipulated in advance, so the scalability is not strong. Secondly, this paper proposes an improved multi-target tracking algorithm for individual motion based on Camshift/Kalman. According to this tracking algorithm, a motion model-based algorithm for individual motion anomaly detection is proposed, in which the proposed multi-target tracking algorithm is used to achieve multi-target tracking. Target tracking automatic initialization, target tracking automatic failure processing and other functions. The proposed individual motion anomaly detection algorithm after the use of individual motion multi-object tracking algorithm to obtain the information of moving targets. Take the smallest block in the video surveillance image as the unit, through the establishment of the normal motion model to detect the unknown moving object, realize the position and the movement direction of the child in the video surveillance. The algorithm has higher accuracy and faster detection speed, especially for individual motion anomaly detection in the scene with fewer moving targets. The average recall rate is 92.6 and the average detection speed is 33.8 Ms / frame, which is better than the experimental data of the same individual anomaly detection algorithm. Third. In this paper, an algorithm for detecting sudden events in children's movement state is proposed, which combines background difference and optical flow method. This algorithm combines the advantages of fast data processing of background difference algorithm and high accuracy of experimental results of optical flow algorithm. For UMN video database, the AUC value of video fragments can reach above 0.96, and the average detection speed is 25.8 Ms / frame. Compared with other similar groups of anomaly detection algorithm in the comprehensive performance is significantly improved. Finally, using the population density estimation method based on pixel statistics, the above proposed security detection algorithms are integrated together. A relatively complete set of video surveillance security detection algorithm suitable for intelligent kindergarten, and the use of OpenCV open source computer vision library. Using Visual C in Windows operating system Microsoft Visual Studio 2010. The key technology of security detection of intelligent kindergarten video surveillance is implemented in this paper.
【學(xué)位授予單位】：浙江大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2017
【分類號(hào)】：TN948.6

【參考文獻(xiàn)】

相關(guān)期刊論文 前10條

1 何傳陽;王平;張曉華;宋丹妮;;基于智能監(jiān)控的中小人群異常行為檢測(cè)[J];計(jì)算機(jī)應(yīng)用;2016年06期

2 唐云;;基于物聯(lián)網(wǎng)技術(shù)構(gòu)建幼兒園安全管理系統(tǒng)[J];信息與電腦(理論版);2016年03期

3 李艷萍;林建輝;楊寧學(xué);;一種基于SIFT特征光流的運(yùn)動(dòng)目標(biāo)跟蹤算法[J];計(jì)算機(jī)科學(xué);2015年11期

4 張薇;于碩;;數(shù)字圖像處理綜述[J];通訊世界;2015年18期

5 張正炎;;高速公路視頻檢測(cè)技術(shù)應(yīng)用[J];道路交通管理;2015年06期

6 羅貴雄;李蕤;章偉;;“智慧校園”的構(gòu)建和應(yīng)用——以深圳市螺嶺外國語實(shí)驗(yàn)學(xué)校為例[J];教育信息技術(shù);2014年12期

7 彭雪莊;陸以勤;鄧健鴻;;智慧校園——學(xué)校教育信息化縱深發(fā)展[J];教育信息技術(shù);2013年12期

8 饒彥;;幼兒園智能視頻監(jiān)控系統(tǒng)探究[J];計(jì)算機(jī)光盤軟件與應(yīng)用;2013年13期

9 王運(yùn)武;;“數(shù)字校園”向“智慧校園”的轉(zhuǎn)型發(fā)展研究——基于系統(tǒng)思維的分析思辯視角[J];遠(yuǎn)程教育雜志;2013年02期

10 �；郾�;;構(gòu)建基于企業(yè)專網(wǎng)的幼兒園安全管理系統(tǒng)[J];計(jì)算機(jī)光盤軟件與應(yīng)用;2013年07期

相關(guān)博士學(xué)位論文 前1條

1 尹宏鵬;基于計(jì)算機(jī)視覺的運(yùn)動(dòng)目標(biāo)跟蹤算法研究[D];重慶大學(xué);2009年

相關(guān)碩士學(xué)位論文 前10條

1 薛璐晨;基于OpenCV的行人異常檢測(cè)系統(tǒng)的設(shè)計(jì)與實(shí)現(xiàn)[D];吉林大學(xué);2016年

2 章強(qiáng);基于視頻檢測(cè)技術(shù)的道路交通參數(shù)監(jiān)測(cè)系統(tǒng)[D];浙江大學(xué);2016年

3 尚進(jìn);監(jiān)控視頻中行人異常行為檢測(cè)系統(tǒng)的設(shè)計(jì)與實(shí)現(xiàn)[D];東南大學(xué);2015年

4 林沁;視頻中的大規(guī)模人群密度與異常行為分析[D];廈門大學(xué);2014年

5 李曉英;基于SSH框架的幼兒園數(shù)字化網(wǎng)絡(luò)管理平臺(tái)設(shè)計(jì)與實(shí)現(xiàn)[D];武漢理工大學(xué);2014年

6 彭懷亮;視頻監(jiān)控場(chǎng)景中人群異常行為識(shí)別研究[D];中國計(jì)量學(xué)院;2014年

7 曹藝華;面向視頻監(jiān)控的大尺度人群行為分析[D];上海交通大學(xué);2013年

8 孫靜;校園網(wǎng)視頻監(jiān)控安全系統(tǒng)[D];天津大學(xué);2012年

9 左常玲;基于視頻的自動(dòng)摔倒檢測(cè)研究與實(shí)現(xiàn)[D];安徽大學(xué);2012年

10 王喬;運(yùn)動(dòng)人體的異常行為檢測(cè)研究[D];電子科技大學(xué);2012年

，

本文編號(hào)：1488607