本文關(guān)鍵詞： 壓縮跟蹤 特征在線選擇 協(xié)方差矩陣 粒子濾波 樣本加權(quán)　出處：《中國(guó)民航大學(xué)》2017年碩士論文　論文類(lèi)型：學(xué)位論文

【摘要】：目標(biāo)跟蹤作為計(jì)算機(jī)視覺(jué)領(lǐng)域的研究熱點(diǎn)與難點(diǎn)已在智能人機(jī)交互、視覺(jué)導(dǎo)航、智能視頻監(jiān)控等領(lǐng)域得到了廣泛的應(yīng)用。但是,研制出一種在各種復(fù)雜場(chǎng)景下均能夠?qū)崿F(xiàn)穩(wěn)定、快速、高精度跟蹤的目標(biāo)跟蹤算法仍然是一項(xiàng)極具挑戰(zhàn)性的課題。近年來(lái),壓縮跟蹤算法因其良好的跟蹤性能成為了研究熱點(diǎn)。為了提高復(fù)雜場(chǎng)景下壓縮跟蹤算法的跟蹤性能,對(duì)其進(jìn)行了研究,主要成果如下:為了提高壓縮跟蹤算法在光照變化、遮擋場(chǎng)景下的跟蹤穩(wěn)定性和準(zhǔn)確性,提出一種結(jié)合特征在線選擇與協(xié)方差矩陣的壓縮跟蹤算法。首先,在特征提取階段引入基于Hellinger距離的特征在線選擇方法,在特征池中動(dòng)態(tài)選擇那些置信水平較高的特征用于構(gòu)建樸素貝葉斯分類(lèi)器。然后,在壓縮跟蹤的框架下融合協(xié)方差矩陣以增強(qiáng)算法對(duì)目標(biāo)的表達(dá)能力,把Haar-like特征和協(xié)方差矩陣相結(jié)合構(gòu)建目標(biāo)模型,取最大響應(yīng)值所對(duì)應(yīng)的候選樣本作為跟蹤結(jié)果。最后,優(yōu)化分類(lèi)器參數(shù)的更新方式,根據(jù)目標(biāo)模板與跟蹤結(jié)果的相似度來(lái)自適應(yīng)地更新分類(lèi)器參數(shù)。針對(duì)壓縮跟蹤算法無(wú)法適應(yīng)目標(biāo)尺度的變化以及沒(méi)有考慮樣本權(quán)重的問(wèn)題,提出一種基于粒子濾波與樣本加權(quán)的壓縮跟蹤算法。首先,對(duì)原始?jí)嚎s跟蹤算法中的壓縮特征進(jìn)行改進(jìn),提取歸一化矩形特征用于構(gòu)建目標(biāo)表觀模型。然后,引入樣本加權(quán)的思想,根據(jù)正樣本與目標(biāo)之間距離的不同賦予正樣本不同的權(quán)重,提高分類(lèi)器的分類(lèi)精度。最后,在粒子濾波的框架下融合尺度不變壓縮特征進(jìn)行動(dòng)態(tài)狀態(tài)估計(jì),在粒子預(yù)測(cè)階段利用一個(gè)二階自回歸模型對(duì)粒子狀態(tài)進(jìn)行估計(jì)與預(yù)測(cè),借助觀測(cè)模型對(duì)粒子狀態(tài)進(jìn)行更新,并且對(duì)粒子進(jìn)行重采樣以防止粒子退化。實(shí)驗(yàn)結(jié)果表明,相比于原始?jí)嚎s跟蹤算法,結(jié)合特征在線選擇與協(xié)方差矩陣的壓縮跟蹤算法在光照變化和遮擋場(chǎng)景下具有更高的跟蹤魯棒性和準(zhǔn)確性;基于粒子濾波與樣本加權(quán)的壓縮跟蹤算法能夠更好的跟蹤目標(biāo)尺度的變化,提高了的跟蹤穩(wěn)定性和準(zhǔn)確性。
[Abstract]:As a hot and difficult point in the field of computer vision, target tracking has been widely used in the fields of intelligent human-computer interaction, visual navigation, intelligent video surveillance and so on. Fast and high precision target tracking algorithm is still a challenging subject. In recent years, compression tracking algorithm has become a research hotspot for its good tracking performance. The main results are as follows: in order to improve the tracking stability and accuracy of the compression tracking algorithm under the illumination variation and occlusion scene, a compression tracking algorithm combining online feature selection and covariance matrix is proposed. In the stage of feature extraction, an online feature selection method based on Hellinger distance is introduced, and those features with high confidence level are dynamically selected in the feature pool to construct naive Bayes classifier. In the framework of compressed tracking, the covariance matrix is fused to enhance the ability of the algorithm to express the target. The target model is constructed by combining the Haar-like features with the covariance matrix, and the candidate samples corresponding to the maximum response value are taken as the tracking results. According to the similarity between the target template and the tracking result, the classifier parameters are updated adaptively. The compression tracking algorithm can not adapt to the change of the target scale and does not consider the weight of the sample. A compression tracking algorithm based on particle filter and sample weighting is proposed. Firstly, the compression feature of the original compression tracking algorithm is improved, and the normalized rectangular feature is extracted to construct the target apparent model. The idea of weighted samples is introduced to give different weights to positive samples according to the distance between positive samples and targets. Finally, the classification accuracy of the classifier is improved. In the framework of particle filter, the dynamic state estimation is carried out by integrating the scale-invariant compression features. In the phase of particle prediction, a second-order autoregressive model is used to estimate and predict the particle state, and the observation model is used to update the particle state. The experimental results show that, compared with the original compression tracking algorithm, the proposed algorithm is better than the original compression tracking algorithm. The compression tracking algorithm based on feature online selection and covariance matrix has higher tracking robustness and accuracy under varying illumination and shading scenes. The compressed tracking algorithm based on particle filter and sample weighting can better track the change of target scale and improve the tracking stability and accuracy.
【學(xué)位授予單位】：中國(guó)民航大學(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10條

1 鄭超;陳杰;殷松峰;楊星;馮云松;凌永順;;改進(jìn)的協(xié)同訓(xùn)練框架下壓縮跟蹤[J];電子與信息學(xué)報(bào);2016年07期

2 李慶武;朱國(guó)慶;周妍;霍冠英;;基于特征在線選擇的目標(biāo)壓縮跟蹤算法[J];自動(dòng)化學(xué)報(bào);2015年11期

3 鐘權(quán);周進(jìn);崔雄文;;融合SIFT特征的壓縮跟蹤算法[J];光電工程;2015年02期

4 盧維;項(xiàng)志宇;于海濱;劉濟(jì)林;;基于自適應(yīng)多特征表觀模型的目標(biāo)壓縮跟蹤[J];浙江大學(xué)學(xué)報(bào)(工學(xué)版);2014年12期

5 張雷;王延杰;何舒文;;基于相位一致性的實(shí)時(shí)壓縮跟蹤方法[J];光子學(xué)報(bào);2014年08期

6 石武禎;寧紀(jì)鋒;顏永豐;;壓縮感知跟蹤中的特征選擇與目標(biāo)模型更新[J];中國(guó)圖象圖形學(xué)報(bào);2014年06期

7 羅會(huì)蘭;鐘寶康;孔繁勝;;結(jié)合目標(biāo)預(yù)測(cè)位置的壓縮跟蹤[J];中國(guó)圖象圖形學(xué)報(bào);2014年06期

8 高文;朱明;賀柏根;吳笑天;;目標(biāo)跟蹤技術(shù)綜述[J];中國(guó)光學(xué);2014年03期

9 邵楓;姜求平;蔣剛毅;郁梅;;基于顯著性分析的立體圖像視覺(jué)舒適度預(yù)測(cè)[J];光學(xué)精密工程;2014年06期

10 鐘權(quán);周進(jìn);吳欽章;王輝;雷濤;;一種改進(jìn)的實(shí)時(shí)壓縮跟蹤算法[J];光電工程;2014年04期

相關(guān)博士學(xué)位論文 前3條

1 陳東成;基于機(jī)器學(xué)習(xí)的目標(biāo)跟蹤技術(shù)研究[D];中國(guó)科學(xué)院研究生院(長(zhǎng)春光學(xué)精密機(jī)械與物理研究所);2015年

2 高峗;復(fù)雜場(chǎng)景下高效視頻目標(biāo)跟蹤關(guān)鍵技術(shù)研究[D];云南大學(xué);2014年

3 鐘權(quán);地空背景下擴(kuò)展目標(biāo)穩(wěn)定跟蹤技術(shù)研究[D];中國(guó)科學(xué)院研究生院（光電技術(shù)研究所）;2014年

相關(guān)碩士學(xué)位論文 前2條

1 朱周元;壓縮跟蹤算法的研究與改進(jìn)[D];安徽大學(xué);2015年

2 李廣三;稀疏表觀模型及高效分類(lèi)器研究[D];沈陽(yáng)理工大學(xué);2014年

，

本文編號(hào)：1501657