本文關(guān)鍵詞： 目標(biāo)跟蹤 模型更新 卷積網(wǎng)絡(luò) 歸一化加權(quán) 多模板　出處：《昆明理工大學(xué)》2017年碩士論文　論文類型：學(xué)位論文

【摘要】：目標(biāo)跟蹤是計(jì)算機(jī)視覺研究領(lǐng)域的熱點(diǎn)之一,本文首先介紹了目標(biāo)跟蹤技術(shù)的研究背景和發(fā)展現(xiàn)狀,描述了目標(biāo)跟蹤中特征提取和運(yùn)動(dòng)估計(jì)兩個(gè)關(guān)鍵部分,并簡(jiǎn)要分析了跟蹤過(guò)程可能遇到的難點(diǎn)。大部分目標(biāo)跟蹤算法可分為兩個(gè)步驟:目標(biāo)特征提取和運(yùn)用目標(biāo)特征實(shí)現(xiàn)跟蹤的算法。在復(fù)雜背景或目標(biāo)被局部遮擋時(shí),目標(biāo)特征信息的唯一性和穩(wěn)定性下降,使得跟蹤算法不能準(zhǔn)確的區(qū)分目標(biāo)和背景,導(dǎo)致跟蹤算法失效;目標(biāo)模型更新策略同樣是維系穩(wěn)定跟蹤的重點(diǎn),跟蹤算法中不可缺少的一部分。目前的模型更新方法僅依賴來(lái)于上一幀或最近幀定位到的目標(biāo)信息,跟蹤的歷史信息未充分利用,當(dāng)發(fā)生遮擋和形變后,跟蹤算法不能準(zhǔn)確的重新定位目標(biāo)。針對(duì)上述問題本文提出算法采用了歸一化距離加權(quán)函數(shù)和多模板模型更新策略,歸一化加權(quán)方法通過(guò)目標(biāo)模板像素到模板中心的距離構(gòu)建加權(quán)函數(shù),在復(fù)雜背景時(shí)可對(duì)目標(biāo)特征進(jìn)行增強(qiáng),減少背景對(duì)目標(biāo)特征信息的干擾;多模板模型更新策略可以在跟蹤過(guò)程提供更完備的目標(biāo)模型匹配信息,本文基于該更新模型結(jié)合卷積網(wǎng)絡(luò)提出一種新的運(yùn)動(dòng)目標(biāo)跟蹤方法。與目前熱點(diǎn)運(yùn)動(dòng)目標(biāo)跟蹤方法在V0T2015運(yùn)動(dòng)目標(biāo)跟蹤測(cè)試視頻集下的對(duì)比實(shí)驗(yàn)表明,本文方法對(duì)于遮擋現(xiàn)象和目標(biāo)自身形變具有較強(qiáng)的魯棒性和較高的準(zhǔn)確性。
[Abstract]:Target tracking is one of the hotspots in the field of computer vision. Firstly, this paper introduces the background and development of target tracking technology, and describes two key parts of feature extraction and motion estimation in target tracking. Most of the target tracking algorithms can be divided into two steps: target feature extraction and target feature tracking algorithm. The uniqueness and stability of the target feature information is decreased, which makes the tracking algorithm unable to distinguish between the target and the background accurately, which leads to the failure of the tracking algorithm. The updating strategy of the target model is also the key to maintain the stable tracking. The current model updating method only depends on the target information located in the previous or most recent frame, and the historical information of the tracking is not fully utilized, when occlusion and deformation occur, The tracking algorithm can not accurately relocate the target. In view of the above problems, this paper proposes a normalized distance weighting function and a multi-template model updating strategy. The normalized weighting method constructs the weighting function through the distance between the target template pixel and the template center, which can enhance the target feature and reduce the interference of the background to the target feature information when the background is complex. Multi-template model updating strategy can provide more complete target model matching information in the tracking process. Based on the updated model and convolutional network, a new moving target tracking method is proposed in this paper, which is compared with the current hot moving target tracking method in V0T2015 moving target tracking video set. The proposed method is robust and accurate to occlusion and target deformation.
【學(xué)位授予單位】：昆明理工大學(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3條

1 李寰宇;畢篤彥;楊源;查宇飛;覃兵;張立朝;;基于深度特征表達(dá)與學(xué)習(xí)的視覺跟蹤算法研究[J];電子與信息學(xué)報(bào);2015年09期

2 沈志熙;楊欣;黃席樾;;均值漂移算法中的目標(biāo)模型更新方法研究[J];自動(dòng)化學(xué)報(bào);2009年05期

3 侯志強(qiáng);韓崇昭;;視覺跟蹤技術(shù)綜述[J];自動(dòng)化學(xué)報(bào);2006年04期

，

本文編號(hào)：1550480