本文關(guān)鍵詞： 行人檢測(cè) 遮擋 HOG 多分類器 BING　出處：《南昌航空大學(xué)》2017年碩士論文　論文類型：學(xué)位論文

【摘要】：隨著人工智能和深度學(xué)習(xí)技術(shù)的不斷發(fā)展,目標(biāo)檢測(cè)、識(shí)別、跟蹤等計(jì)算機(jī)視覺應(yīng)用已經(jīng)越來越多地出現(xiàn)在我們的生活之中。行人檢測(cè)由于其應(yīng)用領(lǐng)域的廣泛,近年來得到了學(xué)者們的廣泛關(guān)注。由于行人自身的非剛性以及所處環(huán)境的復(fù)雜性,各類場(chǎng)景下的行人檢測(cè)依舊是目標(biāo)檢測(cè)領(lǐng)域的難點(diǎn)與重點(diǎn)。本文首先介紹了行人檢測(cè)領(lǐng)域的研究背景與意義,結(jié)合具體事例闡述了行人檢測(cè)在實(shí)際生活中的各項(xiàng)應(yīng)用。然后介紹了目前行人檢測(cè)領(lǐng)域的技術(shù)難點(diǎn)與國(guó)內(nèi)外研究現(xiàn)狀,并對(duì)常用的特征算子與分類方法做了詳細(xì)的介紹。本文的核心工作是針對(duì)日常街道場(chǎng)景中各類不同遮擋狀況的行人,提出一種多分類器級(jí)聯(lián)檢測(cè)的方式,并通過訓(xùn)練通用行人目標(biāo)檢測(cè)BING(binarized normed gradients)模板代替滑動(dòng)窗口掃描來加速檢測(cè)過程。本文的主要工作可以分為以下五個(gè)方面:1、基于HOG(histogram of oriented gradient)特征和線性SVM(Support Vector Machines)分類器,從INRIA數(shù)據(jù)集挑選合適的正負(fù)樣本訓(xùn)練了一個(gè)無遮擋行人分類器。該分類器用于研究行人不同部位被遮擋對(duì)檢測(cè)效果的影響是否不同。驗(yàn)證實(shí)驗(yàn)結(jié)果的數(shù)據(jù)一部分來自CVC-05 part occlusion數(shù)據(jù)集,一部分來自自己收集制作的遮擋行人數(shù)據(jù)集。2、基于INRIA、CVC05-PartOcclusion數(shù)據(jù)集和自己收集的行人圖片制作了一系列遮擋行人專用正樣本,并利用這些數(shù)據(jù)集訓(xùn)練了九個(gè)檢測(cè)不同部位被遮擋的行人分類器。其中無遮擋行人分類器一個(gè),腿部遮擋行人分類器兩個(gè),左右半身遮擋行人分類器各三個(gè)。3、通過級(jí)聯(lián)的方式加速多分類器檢測(cè)過程。分類器設(shè)置為兩級(jí),其中第一級(jí)一個(gè)分類器,第二級(jí)八個(gè)分類器,兩級(jí)分類器之間串聯(lián),第二級(jí)分類器之間并聯(lián)。只有檢測(cè)得分大于第一級(jí)分類器設(shè)置的閾值的檢測(cè)窗口才會(huì)輸入到第二級(jí)分類器做進(jìn)一步的檢測(cè)。通過級(jí)聯(lián)分類器的方式使得檢測(cè)時(shí)間與單分類器基本持平。4、在多分類器檢測(cè)結(jié)果融合階段提出一種NMS+Merging的方式。首先對(duì)于同一分類器的所有檢測(cè)窗口采取NMS的方式去除得分低的檢測(cè)窗口,然后對(duì)于不同分類器保留下來的檢測(cè)窗口采取一種兩兩融合的方式進(jìn)而得到最終的檢測(cè)結(jié)果。5、針對(duì)目標(biāo)檢測(cè)領(lǐng)域滑動(dòng)窗口檢測(cè)法提取候選窗口數(shù)量冗余的問題提出一種基于BING算法改進(jìn)的通用行人目標(biāo)快速提取算法。采用Caltech數(shù)據(jù)集訓(xùn)練,并根據(jù)行人特殊寬高比設(shè)置行人檢測(cè)專用BING模板。該方法可以在保持原有檢測(cè)精度的同時(shí)進(jìn)一步縮短檢測(cè)階段所消耗的時(shí)間。
[Abstract]:With the development of artificial intelligence and depth learning technology, the application of computer vision, such as target detection, recognition, tracking and so on, has been more and more popular in our life. In recent years, scholars have paid more and more attention to it. Due to the nonrigid nature of pedestrians and the complexity of their environment, Pedestrian detection in all kinds of scenarios is still a difficult and important point in the field of target detection. Firstly, this paper introduces the background and significance of pedestrian detection. In this paper, the application of pedestrian detection in real life is expounded with concrete examples. Then, the technical difficulties in the field of pedestrian detection and the current research situation at home and abroad are introduced. The main work of this paper is to propose a multi-classifier cascade detection method for pedestrians with different occlusion conditions in street scenes. The detection process is accelerated by training the general pedestrian target detection BING(binarized normed gradientstemplate instead of sliding window scanning. The main work of this paper can be divided into the following five aspects: 1, based on HOG(histogram of oriented gradient-based feature and linear SVM(Support Vector machines classifier. An unoccluded pedestrian classifier is trained by selecting the appropriate positive and negative samples from the INRIA data set. The classifier is used to study whether the influence of different parts of the pedestrian on the detection effect is different. From the CVC-05 part occlusion dataset, Some of them are collected and made by themselves. 2. Based on the INRIAA CVC05-PartOcclusion dataset and the pedestrian images collected by them, a series of special positive samples of occluded pedestrians are made. Using these data sets, nine pedestrian classifiers are trained to detect the occlusion of different parts, including one unoccluded pedestrian classifier and two leg occluded pedestrian classifiers. The pedestrian classifiers, three from the left and the other half, speed up the process of multi-classifier detection by cascading. The classifier is set to two levels, one classifier in the first stage, eight classifiers in the second stage, and the other in series between the two classifiers. Only a detection window with a detection score greater than the threshold set by the first level classifier will be input to the second level classifier for further detection. The detection time is made by cascading classifiers. At the fusion stage of multi-classifier detection results, a NMS Merging method is proposed. Firstly, NMS is used to remove the low-score detection window for all detection windows of the same classifier. Then, for the detection window retained by different classifiers, a pairwise fusion method is adopted to obtain the final detection result .5. the problem of extracting redundant number of candidate windows by sliding window detection method in target detection field is raised. This paper presents an improved fast pedestrian target extraction algorithm based on BING algorithm. Caltech data set is used to train the pedestrian target. According to the special aspect ratio of pedestrians, a special pedestrian detection BING template is set up. This method can further shorten the time consumed in the detection stage while maintaining the original detection accuracy.
【學(xué)位授予單位】：南昌航空大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2017
【分類號(hào)】：TP391.41

【參考文獻(xiàn)】

相關(guān)期刊論文 前2條

1 曾接賢;程瀟;;結(jié)合單雙行人DPM模型的交通場(chǎng)景行人檢測(cè)[J];電子學(xué)報(bào);2016年11期

2 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)碩士學(xué)位論文 前4條

1 張強(qiáng);靜態(tài)圖像上的行人檢測(cè)方法研究[D];中國(guó)科學(xué)技術(shù)大學(xué);2015年

2 王斌;基于深度學(xué)習(xí)的行人檢測(cè)[D];北京交通大學(xué);2015年

3 姚棲;基于車載攝像頭的行人檢測(cè)與跟蹤方法研究[D];北京理工大學(xué);2015年

4 陳金輝;靜態(tài)圖像行人檢測(cè)算法研究[D];華東理工大學(xué);2015年

，

本文編號(hào)：1540231