本文選題：智能前端　切入點(diǎn)：行人檢測(cè)　出處：《大連海事大學(xué)》2017年碩士論文

【摘要】：隨著政治、經(jīng)濟(jì)的發(fā)展,各個(gè)國(guó)家、企業(yè)、個(gè)人越來(lái)越關(guān)注安防事業(yè)。監(jiān)控系統(tǒng)由最原始的模擬視頻和人眼監(jiān)測(cè)到中期的半數(shù)字化存儲(chǔ)再到如今的全數(shù)字化監(jiān)控系統(tǒng),互聯(lián)網(wǎng)的發(fā)展、編解碼算法的升級(jí)都功不可沒(méi)。而智能前端監(jiān)控系統(tǒng)在監(jiān)控系統(tǒng)中脫穎而出,算法的多樣性需求和前端處理器的飛躍發(fā)展,使得智能前端監(jiān)控系統(tǒng)的廣泛應(yīng)用成為必然。多年來(lái),行人檢測(cè)課題的研究持續(xù)不斷。行人檢測(cè)算法在電子卡口、無(wú)人車(chē)行人避讓系統(tǒng)、客流量檢測(cè)等應(yīng)用中作為基礎(chǔ)算法有著至關(guān)重要的作用。行人檢測(cè)智能前端是帶有行人檢測(cè)分析功能的智能前端,不但能夠代替人眼進(jìn)行監(jiān)查,而且能夠減少傳輸信號(hào)所占用的帶寬和存儲(chǔ)資源。本文根據(jù)《安防監(jiān)控視頻實(shí)時(shí)智能分析設(shè)備技術(shù)要求》設(shè)計(jì)了行人檢測(cè)智能前端系統(tǒng)的功能和性能要求。通過(guò)分析DM8127的優(yōu)勢(shì),確定以DM8127為主處理器的網(wǎng)絡(luò)攝像機(jī)作為系統(tǒng)實(shí)現(xiàn)的硬件平臺(tái),并分析了行人檢測(cè)智能前端的五個(gè)模塊,同時(shí)結(jié)合系統(tǒng)的軟硬件平臺(tái),選用支持向量機(jī)(Support Vector Machine,SVM)和梯度方向直方圖(Histogram of Oriented Gradient,HOG)相結(jié)合的方法作為前端分析模塊的實(shí)現(xiàn)方案。在MATLAB上模擬了行人檢測(cè)系統(tǒng),包括提取HOG特征模塊、圖像金字塔檢測(cè)模塊以及多窗口融合模塊。根據(jù)智能前端行人檢測(cè)的實(shí)時(shí)性和準(zhǔn)確率要求,針對(duì)HOG特征的三個(gè)缺點(diǎn)給出相應(yīng)的解決方法:1)HOG特征的縮放不變性差。選擇包含不同高度行人的同尺寸圖像數(shù)據(jù)集,通過(guò)圖像金字塔檢測(cè)原理,設(shè)計(jì)三層圖像金字塔,并在各層進(jìn)行行人檢測(cè),分析HOG特征縮放不變性性能,結(jié)論為64*128大小的檢測(cè)窗口可以檢測(cè)到行人高度范圍在88像素-128像素內(nèi)的行人。根據(jù)此結(jié)論給出了單層圖像金字塔檢測(cè)法。2)HOG特征的特征維度較高。特征維度過(guò)高導(dǎo)致提取特征耗時(shí)長(zhǎng),檢測(cè)速度緩慢,可在保證準(zhǔn)確率的前提下進(jìn)行算法參數(shù)的適當(dāng)調(diào)整。3)HOG特征對(duì)被遮擋的行人,檢測(cè)效果較差。挑選只有上半身包含頭肩信息的行人作為部分訓(xùn)練正樣本。檢測(cè)到行人后,根據(jù)坐標(biāo)在界面顯示模塊將行人框出,在多尺寸窗口融合技術(shù)的原理上,給出了遞歸的窗口融合算法。仿真后,選用LIBSVM移植到DM8127上用于行人判別,并移植提取HOG特征的算法,實(shí)現(xiàn)整個(gè)智能監(jiān)控系統(tǒng)并進(jìn)行測(cè)試。實(shí)測(cè)結(jié)果表明:添加只包含頭肩信息的行人做為訓(xùn)練正樣本,可以有效地解決行人下半身被遮擋的問(wèn)題;通過(guò)調(diào)整HOG特征的提取參數(shù),可以在保證精度符合要求的情況下,有效提高檢測(cè)速度;給出的遞歸的窗口融合算法,可以有效地將多個(gè)窗口融合;HOG和SVM相結(jié)合的算法移植到DM8127中可以檢測(cè)到90%以上的行人。
[Abstract]:With the development of politics and economy, various countries, enterprises and individuals are paying more and more attention to the security cause. The monitoring system is from the most primitive analog video and the human eye to the semi-digital storage in the medium term and then to the full-digital monitoring system now. The development of the Internet, the upgrading of coding and decoding algorithms, and intelligent front-end monitoring system stand out in the monitoring system, the diversity of algorithms and the rapid development of front-end processors, It makes the wide application of intelligent front-end monitoring system inevitable. Over the years, the research on pedestrian detection has continued. Pedestrian detection algorithm in electronic bayonet, unmanned pedestrian avoidance system, The intelligent front end of pedestrian detection is an intelligent front end with the function of pedestrian detection and analysis. Moreover, it can reduce the bandwidth and storage resources of transmission signal. According to the Technical requirements of Real-time Intelligent Analysis equipment for Security Surveillance Video, this paper designs the function and performance requirements of intelligent front-end system for pedestrian detection. By analyzing the advantages of DM8127, this paper analyzes the advantages of this system. The network camera with DM8127 as the main processor is determined as the hardware platform of the system, and the five modules of the intelligent front end of pedestrian detection are analyzed. At the same time, the hardware and software platform of the system is combined with the hardware and software platform of the system. Support vector machine (SVM) and gradient histogram of Oriented histogram (histogram of Oriented gradient histogram) are selected as the implementation of the front-end analysis module. The pedestrian detection system is simulated on MATLAB, including extracting the HOG feature module. Image pyramid detection module and multi-window fusion module. According to the real-time and accuracy requirements of intelligent front-end pedestrian detection, Aiming at the three shortcomings of HOG feature, this paper gives the corresponding solution, that is, the scaling invariance difference of the HOG feature. The same size image data set including different height pedestrians is selected, and the three-layer image pyramid is designed by the principle of image pyramid detection. At the same time, pedestrian detection is carried out in each layer, and the scaling invariance of HOG features is analyzed. Conclusion the detection window with the size of 64m 128 can detect pedestrians whose height ranges from 88 pixels to 128 pixels. According to this conclusion, the feature dimension of pyramid detection method of single-layer image is higher and the characteristic dimension is too high. It takes a long time to extract the feature, The detection speed is slow, and the algorithm parameters can be adjusted properly on the premise of ensuring the accuracy. The detection effect is poor. Only the upper half of the pedestrian with head-shoulder information as part of the training positive sample. After detecting the pedestrian, according to the coordinates display module in the interface to frame the pedestrian, in the principle of multi-size window fusion technology, A recursive window fusion algorithm is presented. After simulation, LIBSVM is transplanted to DM8127 for pedestrian discrimination, and the algorithm for extracting HOG features is transplanted. The whole intelligent monitoring system is implemented and tested. The experimental results show that adding the pedestrian with only head-shoulder information as the training positive sample can effectively solve the problem that the lower body of the pedestrian is occluded, and adjust the extraction parameters of the HOG feature. The proposed recursive window fusion algorithm can effectively transplant the algorithm combining multiple windows with hog and SVM into DM8127 to detect more than 90% of pedestrians.
【學(xué)位授予單位】：大連海事大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2017
【分類(lèi)號(hào)】：TP391.41

【參考文獻(xiàn)】

相關(guān)期刊論文 前6條

1 盧昆鵬;潘宏俠;;移植Libsvm軟件實(shí)現(xiàn)TMS320F28335的支持向量機(jī)[J];單片機(jī)與嵌入式系統(tǒng)應(yīng)用;2016年03期

2 雷林;潘幸子;楊敏;孟麗珍;;HOG特征及其應(yīng)用研究[J];信息通信;2016年01期

3 芮挺;費(fèi)建超;周怞;方虎生;朱經(jīng)緯;;基于深度卷積神經(jīng)網(wǎng)絡(luò)的行人檢測(cè)[J];計(jì)算機(jī)工程與應(yīng)用;2016年13期

4 李梅芳;李輝;;網(wǎng)絡(luò)攝像機(jī)的優(yōu)勢(shì)分析[J];企業(yè)技術(shù)開(kāi)發(fā);2015年18期

5 劉哲夫;;基于DSP平臺(tái)的行人檢測(cè)的實(shí)現(xiàn)和優(yōu)化[J];中國(guó)高新技術(shù)企業(yè);2013年36期

6 蘇松志;李紹滋;陳淑媛;蔡國(guó)榕;吳云東;;行人檢測(cè)技術(shù)綜述[J];電子學(xué)報(bào);2012年04期

相關(guān)會(huì)議論文 前1條

1 柳建為;應(yīng)娜;楊慶彪;;基于HOG特征與多尺度窗口融合的行人檢測(cè)算法[A];信號(hào)處理在生儀2014學(xué)術(shù)年會(huì)論文集[C];2014年

相關(guān)碩士學(xué)位論文 前10條

1 向根;基于DM8127的多目標(biāo)遠(yuǎn)距離檢測(cè)定位系統(tǒng)[D];電子科技大學(xué);2016年

2 楊芬;基于DM6467的智能視頻監(jiān)控前端的設(shè)計(jì)與實(shí)現(xiàn)[D];大連海事大學(xué);2015年

3 崔劍;基于多特征融合的分級(jí)行人檢測(cè)方法研究[D];電子科技大學(xué);2015年

4 劉琳;基于人體頭肩特征的行人檢測(cè)方法研究與應(yīng)用[D];南京理工大學(xué);2015年

5 戴毅;行人檢測(cè)算法及其在DM8168平臺(tái)上的實(shí)現(xiàn)[D];上海交通大學(xué);2015年

6 何諧;基于DSP優(yōu)化的行人識(shí)別算法在智能監(jiān)控中的研究與應(yīng)用[D];電子科技大學(xué);2014年

7 傅智勇;HOG+SVM行人檢測(cè)算法在DM6437上的實(shí)現(xiàn)與優(yōu)化[D];華南理工大學(xué);2012年

8 陳剛;基于多特征的行人檢測(cè)方法研究[D];吉林大學(xué);2012年

9 胡將勝;基于AdaBoost和SVM的人體檢測(cè)[D];中南民族大學(xué);2011年

10 嚴(yán)照宇;基于視頻分析的智能監(jiān)控系統(tǒng)研究與實(shí)現(xiàn)[D];電子科技大學(xué);2010年

，

本文編號(hào)：1674248