本文選題：魚(yú)眼攝像頭 + 全景圖像��； 參考：《中國(guó)農(nóng)業(yè)大學(xué)》2017年博士論文

【摘要】：基于全景視覺(jué)的汽車行駛環(huán)境監(jiān)測(cè)系統(tǒng)可以為駕駛員提供全景成像和目標(biāo)檢測(cè)兩大駕駛輔助功能,是主動(dòng)安全領(lǐng)域的重要技術(shù),具有重要的應(yīng)用價(jià)值�，F(xiàn)有的目標(biāo)檢測(cè)功能主要基于雷達(dá)或普通攝像頭實(shí)現(xiàn),其檢測(cè)算法對(duì)于大視場(chǎng)、大畸變的魚(yú)眼攝像頭并不適用。本文首先對(duì)全景成像系統(tǒng)的標(biāo)定方法進(jìn)行改進(jìn)從而快速獲得精確的系統(tǒng)參數(shù),然后對(duì)基于全景視覺(jué)的行人檢測(cè)算法進(jìn)行研究。主要研究?jī)?nèi)容包括:分析了全景成像系統(tǒng)的標(biāo)定原理。通過(guò)對(duì)比實(shí)驗(yàn)選擇折反射模型作為魚(yú)眼攝像頭的成像模型,以此為基礎(chǔ)進(jìn)行攝像頭的標(biāo)定和全景成像系統(tǒng)的標(biāo)定。為魚(yú)眼攝像頭設(shè)計(jì)一種由三個(gè)相互垂直的標(biāo)定板組成的立體標(biāo)定板,使標(biāo)定時(shí)角點(diǎn)完全覆蓋攝像頭,從而充分利用魚(yú)眼圖像邊緣部分以獲得更精確的攝像頭參數(shù)。在全景系統(tǒng)標(biāo)定板上添加定位塊,保證系統(tǒng)標(biāo)定的穩(wěn)定性。對(duì)魚(yú)眼圖像進(jìn)行前視圖投影和直方圖均衡化,完成預(yù)處理。建立歸一化的行人樣本庫(kù)。分別對(duì)常用的行人描述特征和機(jī)器學(xué)習(xí)算法進(jìn)行分析,包括Haar特征結(jié)合AdaBoost算法,HOG特征結(jié)合SVM算法,以及卷積神經(jīng)網(wǎng)絡(luò)進(jìn)行分類器訓(xùn)練實(shí)驗(yàn),對(duì)所得分類器和訓(xùn)練過(guò)程進(jìn)行評(píng)價(jià),總結(jié)各自的特點(diǎn)。對(duì)行人局部HOG特征與整體HOG特征進(jìn)行對(duì)比,提出基于ROI-HOG特征訓(xùn)練SVM分類器;設(shè)計(jì)卷積神經(jīng)網(wǎng)絡(luò)行人檢測(cè)分類器,優(yōu)化網(wǎng)絡(luò)結(jié)構(gòu)參數(shù),并結(jié)合無(wú)監(jiān)督CNN提取特征與線性SVM分類器監(jiān)督學(xué)習(xí)形成組合分類器,實(shí)現(xiàn)快速準(zhǔn)確的行人檢測(cè)。為充分利用魚(yú)眼圖像的大視場(chǎng),研究如何克服圖像邊緣的大幅度變形。提出了多橫擺角前視投影的方法,將一幅魚(yú)眼圖像展開(kāi)為不同橫擺角的前視投影圖,然后進(jìn)行行人檢測(cè)。圖像任意位置的行人在某一范圍內(nèi)的虛擬橫擺角下的前視圖中,都可去除仿射變形恢復(fù)正常人體比例,方便后續(xù)進(jìn)行行人檢測(cè)。通過(guò)實(shí)驗(yàn)總結(jié)前視圖橫擺角設(shè)置規(guī)則,盡量減少一幅魚(yú)眼圖像的展開(kāi)數(shù),減少檢測(cè)耗時(shí)。進(jìn)行實(shí)車全景系統(tǒng)標(biāo)定實(shí)驗(yàn),結(jié)果表明系統(tǒng)具有較高的標(biāo)定精度和標(biāo)定穩(wěn)定性。采集行車視頻數(shù)據(jù),逐幀標(biāo)記行人形成測(cè)試集,在PC上設(shè)計(jì)評(píng)估軟件,運(yùn)行行人檢測(cè)程序,評(píng)估行人檢測(cè)算法性能,結(jié)果表明在一定距離范圍內(nèi)本文算法可以實(shí)現(xiàn)較高的行人檢測(cè)率。
[Abstract]:The vehicle driving environment monitoring system based on panoramic vision can provide driving assistant functions of panoramic imaging and target detection. It is an important technology in active safety field and has important application value. The existing target detection function is mainly based on radar or ordinary camera, and its detection algorithm is not suitable for large field of view and large distortion fish-eye camera. In this paper, the calibration method of panoramic imaging system is improved to obtain the accurate system parameters quickly, and then the pedestrian detection algorithm based on panoramic vision is studied. The main contents are as follows: the calibration principle of panoramic imaging system is analyzed. The refraction model is chosen as the imaging model of fish-eye camera by contrast experiment, and the camera and panoramic imaging system are calibrated based on the model. A stereo calibration board composed of three vertical calibration boards is designed for the fish-eye camera, which can cover the camera completely when the corners are calibrated, thus making full use of the edge part of the fish-eye image to obtain more accurate camera parameters. A positioning block is added to the panoramic system calibration board to ensure the stability of the system calibration. The front view projection and histogram equalization are used to preprocess the fish eye image. A normalized pedestrian sample bank is established. The common pedestrian description features and machine learning algorithms are analyzed respectively, including Haar feature and AdaBoost algorithm combined with SVM algorithm, and convolution neural network for classifier training experiment. The classifier and training process are evaluated. Summarize their own characteristics. By comparing the local HOG features of pedestrians with the overall HOG features, a SVM classifier based on ROI-HOG feature training is proposed, and a pedestrian detection classifier based on convolution neural network is designed to optimize the network structure parameters. Combined with unsupervised CNN feature extraction and linear SVM classifier supervised learning, a combined classifier is formed to realize fast and accurate pedestrian detection. In order to make full use of the large field of view of the fish-eye image, this paper studies how to overcome the large deformation of the image edge. A method of forward projection with multiple yaw angles is proposed. A fish-eye image is expanded into a forward projection image with different yaw angles, and then pedestrian detection is carried out. In the front view of a virtual yaw angle in any position of the image, the affine deformation can be removed to restore the normal proportion of the human body, and it is convenient to carry out pedestrian detection. The rules of yaw angle setting in front view are summarized by experiments to reduce the expansion number of a fish-eye image and the detection time. The calibration experiment of real vehicle panoramic system shows that the system has high calibration accuracy and stability. Collect the video data of the vehicle, mark the pedestrian to form the test set frame by frame, design the evaluation software on the PC, run the pedestrian detection program, evaluate the performance of the pedestrian detection algorithm, The results show that the proposed algorithm can achieve high pedestrian detection rate within a certain range of distances.
【學(xué)位授予單位】：中國(guó)農(nóng)業(yè)大學(xué)
【學(xué)位級(jí)別】：博士
【學(xué)位授予年份】：2017
【分類號(hào)】：U463.6

【參考文獻(xiàn)】

相關(guān)期刊論文 前10條

1 高秀麗;遲瑞娟;王建強(qiáng);;采用立體標(biāo)定板的魚(yú)眼相機(jī)快速標(biāo)定方法[J];哈爾濱工程大學(xué)學(xué)報(bào);2016年11期

2 孫銳;張廣海;高雋;;基于深度分層特征表示的行人識(shí)別方法[J];電子與信息學(xué)報(bào);2016年06期

3 曾敏;周益龍;;基于深度學(xué)習(xí)模型的行人檢測(cè)研究與仿真[J];南京郵電大學(xué)學(xué)報(bào)(自然科學(xué)版);2015年06期

4 丁文秀;孫銳;閆曉星;;基于分層深度學(xué)習(xí)的魯棒行人分類[J];光電工程;2015年09期

5 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

6 嚴(yán)春燕;;全景攝像機(jī)標(biāo)定方法綜述[J];電腦知識(shí)與技術(shù);2014年34期

7 遲瑞娟;趙木禎;王建強(qiáng);張德兆;;一種基于折反射相機(jī)的客車全景環(huán)視系統(tǒng)[J];哈爾濱工程大學(xué)學(xué)報(bào);2013年02期

8 江鐵;朱桂斌;孫奧;;全景圖像拼接技術(shù)研究現(xiàn)狀綜述[J];重慶工商大學(xué)學(xué)報(bào)(自然科學(xué)版);2012年12期

9 孫志軍;薛磊;許陽(yáng)明;王正;;深度學(xué)習(xí)研究綜述[J];計(jì)算機(jī)應(yīng)用研究;2012年08期

10 郭立君;劉曦;趙杰煜;史忠植;;結(jié)合運(yùn)動(dòng)信息與表觀特征的行人檢測(cè)方法[J];軟件學(xué)報(bào);2012年02期

相關(guān)博士學(xué)位論文 前5條

1 賴世銘;全景凝視系統(tǒng)中的關(guān)鍵技術(shù)研究[D];國(guó)防科學(xué)技術(shù)大學(xué);2014年

2 張陽(yáng);復(fù)雜交通場(chǎng)景中基于視頻的行人檢測(cè)與跟蹤若干關(guān)鍵問(wèn)題研究[D];華南理工大學(xué);2013年

3 馮為嘉;基于魚(yú)眼鏡頭的全方位視覺(jué)及全景立體球視覺(jué)研究[D];天津大學(xué);2012年

4 汪中;面向變化場(chǎng)景的行人分類檢測(cè)方法研究[D];中國(guó)科學(xué)技術(shù)大學(xué);2011年

5 郭烈;基于單目視覺(jué)的車輛前方行人檢測(cè)技術(shù)研究[D];吉林大學(xué);2007年

，

本文編號(hào)：1862021