本文選題：計(jì)算機(jī)視覺　切入點(diǎn)：智能視頻系統(tǒng)　出處：《電子科技大學(xué)》2016年博士論文

【摘要】：基于視頻的目標(biāo)跟蹤與識(shí)別技術(shù)是計(jì)算機(jī)視覺的主要研究方向之一,是諸如智能監(jiān)控、人機(jī)交互、地形導(dǎo)航及視頻智能標(biāo)注檢索等應(yīng)用的基礎(chǔ)和關(guān)鍵技術(shù),是實(shí)現(xiàn)“智慧城市”、“平安城市”的重要手段,具有重要的理論研究與實(shí)際應(yīng)用價(jià)值。然而,自然非受控條件下獲取的視頻中,環(huán)境復(fù)雜多變,對(duì)其中各類目標(biāo)跟蹤識(shí)別帶來諸多挑戰(zhàn)。針對(duì)各種復(fù)雜場(chǎng)景及不同目標(biāo),如何設(shè)計(jì)實(shí)現(xiàn)效率高、魯棒性好、實(shí)時(shí)性強(qiáng)的目標(biāo)跟蹤識(shí)別技術(shù)仍然是當(dāng)今業(yè)界研究的熱點(diǎn)及難點(diǎn)。鑒于此認(rèn)識(shí),本文立足于前人一系列優(yōu)秀成果,啟發(fā)于人類自身視覺系統(tǒng)完美的目標(biāo)發(fā)現(xiàn)跟蹤與識(shí)別機(jī)制,主要針對(duì)視頻中人體目標(biāo)的跟蹤與識(shí)別問題展開了深入的研究,取得的創(chuàng)新性成果如下:1.針對(duì)目標(biāo)樣本稀缺,致使目標(biāo)特征初始不充分,抑或目標(biāo)原有特征易被遮擋、偽裝,甚或因時(shí)光荏苒而逐漸滅失消亡,從而導(dǎo)致目標(biāo)已知特征數(shù)據(jù)逐漸失效,如此種種,最終導(dǎo)致目標(biāo)跟蹤識(shí)別效率極低,提出了一種基于單樣本的自主在線的目標(biāo)特征學(xué)習(xí)及更新算法(One Sample Based Autonomous Online Features Learning and Updating Algorithm,FLUA)。算法首先基于目標(biāo)單樣本獲取的局部特征,對(duì)視頻中目標(biāo)的相似視圖區(qū)域進(jìn)行識(shí)別定位,爾后根據(jù)視頻幀之間靜態(tài)特征點(diǎn)的值、分布及其運(yùn)動(dòng)一致性等多重匹配的校驗(yàn),在線學(xué)習(xí)更新目標(biāo)新特征,一旦目標(biāo)新特征得到確認(rèn),立即作用于隨后的目標(biāo)跟蹤識(shí)別及其特征學(xué)習(xí)更新過程。實(shí)驗(yàn)結(jié)果表明,本章提出的FLUA算法無需大量目標(biāo)樣本圖像,無需事先大量訓(xùn)練,縱然在單樣本情形下,特征學(xué)習(xí)獲取效果依然顯著,有效的提高了目標(biāo)跟蹤過程的效率。特征學(xué)習(xí)無需復(fù)雜的迭代求解過程,更新速度快,能夠滿足跟蹤系統(tǒng)實(shí)時(shí)性要求。2.針對(duì)人體目標(biāo)頭部姿態(tài)抑或臉部表情等變化、臉部化妝抑或偽裝等等嚴(yán)重影響目標(biāo)人臉部圖像的跟蹤獲取,從而對(duì)基于人臉的目標(biāo)跟蹤與識(shí)別等應(yīng)用系統(tǒng)產(chǎn)生極為不利的影響,提出了一種基于人體模糊跟蹤的人臉跟蹤獲取算法(Human Body Fuzzy Tracking Based Face Tracking and Capturing Algorithm,B-FTC)。算法首先根據(jù)目標(biāo)身體各肢體部分特征及運(yùn)動(dòng)一致性匹配跟蹤定位目標(biāo)的身體,爾后根據(jù)目標(biāo)頭部與身體的位置及運(yùn)動(dòng)相關(guān)性定位獲取目標(biāo)的臉部圖像。在對(duì)目標(biāo)身體跟蹤識(shí)別的同時(shí),引入了在線特征學(xué)習(xí)更新機(jī)制以應(yīng)對(duì)目標(biāo)的外觀特征的逐步變化。實(shí)驗(yàn)結(jié)果證明,該算法對(duì)目標(biāo)頭部姿態(tài)、鏡頭視角、臉部表情等等變化,以及臉部局部遮擋、化妝、偽裝等等不利因素具有完全的魯棒性,同時(shí)具有極好的臉部跟蹤獲取及歸屬分類效果,在自然監(jiān)控視頻及四川變臉表演視頻中,對(duì)目標(biāo)臉部圖像的跟蹤獲取率都在90%以上,正確率幾近達(dá)到100%。3.針對(duì)監(jiān)控視頻不同于生活攝影,其中人物臉部表情及頭部動(dòng)作較多自然變化,獲取的臉部圖像多以不同視角的‘表情碎片’形式存在,從而導(dǎo)致基于正面或近正面表情平靜的人臉識(shí)別算法失效,本文提出了一套可以相當(dāng)程度免疫于頭部姿態(tài)、表情、光線等諸多變化以及部分遮擋等不利情形下,N:M的video-to-video人臉自動(dòng)識(shí)別算法(Space and Expression Double Weighted based Video-to-Video Face Recognition,SEDW-2VFR)。算法首先根據(jù)臉部碎片特征點(diǎn)的值、分布的雙重匹配及其運(yùn)動(dòng)變換的誤差大小對(duì)基準(zhǔn)視頻中跟蹤獲取的目標(biāo)臉部碎片圖像進(jìn)行區(qū)域及表情的雙重分類,對(duì)基準(zhǔn)目標(biāo)的每一臉部圖像類集進(jìn)行特征投影矩陣的生成及特征的提取,而后對(duì)待測(cè)視頻中跟蹤獲取的目標(biāo)人臉部碎片進(jìn)行在線分權(quán)2D-PCA識(shí)別。實(shí)驗(yàn)表明,該算法對(duì)頭部姿態(tài)及表情等變化具有很強(qiáng)的魯棒性,在自然條件下的生活視頻中,目標(biāo)跟蹤識(shí)別率依然達(dá)到90%以上。4.針對(duì)大多數(shù)現(xiàn)有步態(tài)識(shí)別算法預(yù)設(shè)條件苛刻,其中步態(tài)表示、提取及比對(duì)過程復(fù)雜,計(jì)算量大,識(shí)別效果差,提出了一種基于肢體區(qū)域及步態(tài)周期雙重區(qū)分的步態(tài)特征異步提取,同步分權(quán)融合的2D-PCA步態(tài)識(shí)別算法(Limbs and Gait Period Double Distinguished Feature Asynchronous Extraction and Synchronous Weighted Fusion Based Gait Recognition,FAESWF-GR)。算法首先對(duì)基準(zhǔn)目標(biāo)各肢體部分進(jìn)行異步特征提取,并根據(jù)步態(tài)特征周期的長(zhǎng)度進(jìn)行歸類和‘時(shí)間片’劃分,而后采用2D-PCA算法對(duì)不同周期長(zhǎng)度的步態(tài)特征‘時(shí)間片’子集進(jìn)行特征投影矩陣生成及特征提取,然后對(duì)待識(shí)別目標(biāo)肢體各部分進(jìn)行在線的步態(tài)周期特征提取及時(shí)間片劃分,同時(shí)進(jìn)行同步與分權(quán)相融合的2D-PCA步態(tài)識(shí)別。實(shí)驗(yàn)表明,算法的步態(tài)特征異步提取機(jī)制具備了對(duì)視頻中身體局部碎片圖像進(jìn)行特征提取的能力,從而使算法對(duì)身體的視角、姿態(tài)、焦距的變化及身體的局部遮擋等等都具有了極強(qiáng)的魯棒性。另外,算法的同步與分權(quán)相融合的綜合比對(duì)機(jī)制中肢體各部分特征權(quán)重可調(diào),從而使算法能針對(duì)不同情形對(duì)肢體各部分特征賦予不同權(quán)重,極易體現(xiàn)肢體各部分在不同情形下步態(tài)的整體性及權(quán)重的差異性,很好適應(yīng)自然條件下視頻中目標(biāo)人身體及其所處環(huán)境的各種復(fù)雜變化,取得較高識(shí)別率。
[Abstract]:Target tracking and recognition technology based on video is one of the main research direction of computer vision, such as intelligent monitoring, human-computer interaction, and based retrieval applications such as terrain navigation and intelligent video annotation, is the realization of "smart city", "an important means of safe city, has the important value of theoretical research and practical application however, non natural obtained under controlled conditions in the video, including all kinds of complicated environment, target tracking and Recognition brings many challenges. For a variety of complex scenes and different goals, how to design and realize the high efficiency, good robustness, real-time target tracking and recognition technology is still the hotspot and difficulty in the research of industry. In view of this understanding, this paper is based on previous a series of excellent achievements, inspired by the human visual system perfect target tracking and recognition mechanism found, to video Focuses on tracking and recognition problems in human target, innovative results obtained are as follows: 1. for the samples are scarce, resulting in the initial target feature is not sufficient, or the original features easily obscured, camouflage, or even because of loss of time flies and gradually die, which leads to target known feature data is invalid. So eventually, the target tracking and recognition efficiency is very low, this paper proposes a target feature of single sample independent online learning and updating algorithm based on (One Sample Based Autonomous Online Features Learning and Updating Algorithm, FLUA). The first algorithm based on local feature single sample acquisition, identification of similar video objects in the view area then, according to the video frame between the static feature point value, check the distribution and motion consistency of multiple matching, learning update in line New features, new features of the target once confirmed, immediately for a subsequent target tracking and recognition and feature learning update process. The experimental results show that the proposed FLUA algorithm does not need a large number of target images, without a lot of training, even in a single sample, obtain the feature learning effect is still significant, effectively improve the efficiency of the process of target tracking. The characteristics of learning without iteration process complex, update speed, can meet the requirements of real-time tracking system for.2. human head pose or facial expression changes, facial makeup or camouflage so seriously affected the facial image of the target tracking and acquisition, target tracking and recognition of face the application system based on an extremely negative impact, puts forward a fuzzy tracking human face tracking algorithm based on Human (Body Fuzzy for Track Ing Based Face Tracking and Capturing Algorithm, B-FTC). Firstly, according to the target body characteristics and motion of each limb matching tracking and positioning the target's body, then according to the position and moving target correlation between the head and body of the target face images. In the target tracking and recognition of the body at the same time, the characteristics of online learning update mechanism to gradually change the appearance characteristics on the target. The experimental results show that the algorithm of head pose, camera angle, facial expressions and so on, and the face is partially occluded, makeup, camouflage and so on unfavorable factors has robustness, while having excellent face tracking acquisition and classification effect in natural video surveillance and Sichuan face video performance, the target tracking facial image acquisition rate is above 90%, the correct rate almost reached for 100%.3. Video monitoring is different from life photography, in which characters face and head movements more natural changes, the face image acquisition exist in the different perspectives of the "expression fragments' form, which leads to positive or positive expression in the calm face recognition algorithm based on failure, this paper presents a considerable degree can be immune to the head pose and expression. Light changes and partial occlusions and many other adverse circumstances, video-to-video face automatic recognition algorithm of N:M (Space and Expression Double Weighted based Video-to-Video Face Recognition, SEDW-2VFR). The algorithm first face feature points according to the values of the fragments, double classification distribution of double matching and motion transform error size and the expression of the target region face image fragments get reference in the video, the reference target each face image feature set cast Generation and characteristics of shadow matrix, and then to be obtained in video target tracking facial fragments online decentralization 2D-PCA recognition. Experiments show that the algorithm has strong robustness to the head pose and facial expression, under natural conditions in the live video, target tracking and recognition rate is still more than 90%.4. for most of the existing gait recognition algorithm design conditions, including gait representation, extraction and alignment process is complex, large amount of calculation, poor recognition effect, put forward a step behavior and gait cycle of double limb asynchronous regional differentiation based gait recognition algorithm 2D-PCA extraction, synchronous fusion (Limbs and Gait power Period Double Distinguished Feature Asynchronous Extraction and Synchronous Weighted Fusion Based Gait Recognition, FAESWF-GR). The algorithm first has the reference target each part of the body Asynchronous feature extraction, and classification and the "time slice" divided according to the gait cycle length, and use the 2D-PCA algorithm to the gait feature "on different period length of time slice 'subset feature projection matrix generation and feature extraction, classification features and extract the gait cycle time slice and then treat each part of body target identification online. At the same time the 2D-PCA gait recognition combined with synchronous decentralization. Experimental results show that the mechanism of gait feature extraction algorithm of asynchronous have the ability of feature extraction for video image fragments of body parts on the body, so that the algorithm from the perspective of attitude, and the focal length of the local change of body occlusion has a strong robustness. In addition, the comprehensive comparison mechanism integration and decentralization of the synchronization algorithm of each part of body feature weight can be adjusted, so that the algorithm can not for sympathy The shape gives different weights to each part of the limbs. It can easily reflect the difference of gait's integrity and weight between different parts of the body, so it can adapt to all kinds of complex changes of the target's body and its environment under natural conditions, and achieve a high recognition rate.

【學(xué)位授予單位】：電子科技大學(xué)
【學(xué)位級(jí)別】：博士
【學(xué)位授予年份】：2016
【分類號(hào)】：TP391.41

【參考文獻(xiàn)】

相關(guān)期刊論文 前3條

1 李丹;王洪濤;;基于粒子重采樣濾波算法的紅外圖像消噪[J];紅外技術(shù);2014年03期

2 劉雪超;吳志勇;黃德天;梁敏華;;結(jié)合視覺感知的調(diào)焦窗口構(gòu)建[J];中國(guó)激光;2014年01期

3 羅海波;史澤林;;紅外成像制導(dǎo)技術(shù)發(fā)展現(xiàn)狀與展望[J];紅外與激光工程;2009年04期

，

本文編號(hào)：1675611