基于最大池圖匹配的形變目標(biāo)跟蹤方法
發(fā)布時間：2018-07-23 15:19
【摘要】：隨著大數(shù)據(jù)時代的到來,計算機技術(shù)和網(wǎng)絡(luò)技術(shù)突飛猛進的發(fā)展,計算機視覺技術(shù)成為信息科學(xué)研究領(lǐng)域的重要課題。而作為諸多計算機視覺高層應(yīng)用的基礎(chǔ),視覺跟蹤技術(shù)也越來越受到國內(nèi)外研究者的重視。根據(jù)實際應(yīng)用,視覺跟蹤主要分為兩個大的方向：單目標(biāo)跟蹤和多目標(biāo)跟蹤。雖然研究者在單目標(biāo)跟蹤課題上做了大量的研究,但是目標(biāo)在運動過程中所包含的各種信息以及場景限制并未得到充分挖掘。單目標(biāo)跟蹤過程中,目標(biāo)可能會產(chǎn)生巨大形變或者面臨嚴(yán)重遮擋,此時目標(biāo)的外觀會發(fā)生巨大變化,這種情況下如果繼續(xù)使用傳統(tǒng)的整體框(bounding box)來描述目標(biāo),勢必會濾掉前景目標(biāo)部分或者引入背景噪聲,無法給出精確的目標(biāo)表達。本文針對單目標(biāo)跟蹤進行相應(yīng)研究和探討,就跟蹤過程中出現(xiàn)的關(guān)鍵技術(shù)難題,提出了基于部件的最大池圖匹配的跟蹤方法(Max-pooling Graph matching based Tracker, MGT)。文章的主要內(nèi)容總結(jié)如下：(1)不同于基于目標(biāo)整體模型的算法,本文算法基于目標(biāo)部件模型,采用動態(tài)圖結(jié)構(gòu)表示目標(biāo)部件,即目標(biāo)部件的表象特征(表象信息),以及它們之間的相對位置關(guān)系(結(jié)構(gòu)信息)。對于目標(biāo)搜索區(qū)域,算法基于圖像分割技術(shù)提取出超像素候選目標(biāo)部件建立候選圖,并與建立好的的目標(biāo)圖模型進行匹配。(2)圖匹配策略采用最大池(max-pooling)圖匹配方法,即目標(biāo)圖匹配對中的每一個節(jié)點支持項都只使用候選圖中的最大池支持項,并將其相關(guān)結(jié)構(gòu)一致性分?jǐn)?shù)作為匹配似然度,建立起目標(biāo)圖模型和候選圖之間的部件匹配關(guān)系。在此基礎(chǔ)上得到目標(biāo)位置的置信圖(confidence map),通過采樣可以確定目標(biāo)的最優(yōu)位置。(3)最后,為了避免僅考慮局部目標(biāo)部件的貢獻造成的鑒別力不夠,我們引入了整體目標(biāo)的特征表達參與目標(biāo)位置投票,以提高跟蹤魯棒性。
[Abstract]:With the coming of big data era and the rapid development of computer technology and network technology, computer vision technology has become an important subject in the field of information science research. As the foundation of many high-level applications of computer vision, visual tracking technology has been paid more and more attention by researchers at home and abroad. According to the practical application, visual tracking is divided into two major directions: single target tracking and multi-target tracking. Although researchers have done a lot of research on the subject of single target tracking, all kinds of information and scene constraints contained in the process of target motion have not been fully exploited. In the process of single target tracking, the target may produce huge deformation or face severe occlusion, and the appearance of the target will change greatly. In this case, if we continue to use the traditional global box (bounding box) to describe the target, It will filter out the foreground target or introduce background noise, so it can not express the target accurately. Based on the research and discussion of single target tracking, this paper presents a tracking method based on component maximum pool map matching (Max-pooling Graph matching based Tracker, MGT).) for the key technical problems in the tracking process. The main contents of this paper are summarized as follows: (1) different from the algorithm based on the whole object model, this algorithm is based on the target component model and uses dynamic graph structure to represent the target component. That is the representation of the target component (representation information) and the relative position relationship between them (structure information). For the target search region, the algorithm extracts candidate target components to build candidate images based on image segmentation technology, and matches them with the established target graph model. (2) the maximum pool (max-pooling) graph matching method is used to match the graph matching strategy. In other words, each node support item in the target graph matching pair only uses the maximum pool support item in the candidate graph, and takes the correlation structure consistency score as the matching likelihood degree, and establishes the component matching relationship between the target graph model and the candidate graph. On this basis, the (confidence map), of the target location can be obtained by sampling the optimal position of the target. (3) finally, in order to avoid considering only the contribution of the local target components, the discriminant ability is not enough. In order to improve the tracking robustness, we introduce the feature representation of the whole target to participate in the target location voting.
【學(xué)位授予單位】：合肥工業(yè)大學(xué)
【學(xué)位級別】：碩士
【學(xué)位授予年份】：2016
【分類號】：TP391.41
本文編號：2139816
[Abstract]:With the coming of big data era and the rapid development of computer technology and network technology, computer vision technology has become an important subject in the field of information science research. As the foundation of many high-level applications of computer vision, visual tracking technology has been paid more and more attention by researchers at home and abroad. According to the practical application, visual tracking is divided into two major directions: single target tracking and multi-target tracking. Although researchers have done a lot of research on the subject of single target tracking, all kinds of information and scene constraints contained in the process of target motion have not been fully exploited. In the process of single target tracking, the target may produce huge deformation or face severe occlusion, and the appearance of the target will change greatly. In this case, if we continue to use the traditional global box (bounding box) to describe the target, It will filter out the foreground target or introduce background noise, so it can not express the target accurately. Based on the research and discussion of single target tracking, this paper presents a tracking method based on component maximum pool map matching (Max-pooling Graph matching based Tracker, MGT).) for the key technical problems in the tracking process. The main contents of this paper are summarized as follows: (1) different from the algorithm based on the whole object model, this algorithm is based on the target component model and uses dynamic graph structure to represent the target component. That is the representation of the target component (representation information) and the relative position relationship between them (structure information). For the target search region, the algorithm extracts candidate target components to build candidate images based on image segmentation technology, and matches them with the established target graph model. (2) the maximum pool (max-pooling) graph matching method is used to match the graph matching strategy. In other words, each node support item in the target graph matching pair only uses the maximum pool support item in the candidate graph, and takes the correlation structure consistency score as the matching likelihood degree, and establishes the component matching relationship between the target graph model and the candidate graph. On this basis, the (confidence map), of the target location can be obtained by sampling the optimal position of the target. (3) finally, in order to avoid considering only the contribution of the local target components, the discriminant ability is not enough. In order to improve the tracking robustness, we introduce the feature representation of the whole target to participate in the target location voting.
【學(xué)位授予單位】：合肥工業(yè)大學(xué)
【學(xué)位級別】：碩士
【學(xué)位授予年份】：2016
【分類號】：TP391.41
【相似文獻】
相關(guān)會議論文 前3條
1 金心;劉漢江;王定坤;吳曄;張新文;賀軍華;;枕大池切開引流術(shù)治療擴張性第四腦室積血[A];2009年浙江省神經(jīng)外科學(xué)術(shù)年會論文匯編[C];2009年
2 包長順;楊福兵;;枕大池重建術(shù)治療合并脊髓空洞的Arnold-Chiari畸形[A];2011中華醫(yī)學(xué)會神經(jīng)外科學(xué)學(xué)術(shù)會議論文匯編[C];2011年
3 黃銀洲;王乃昂;何彤慧;隆浩;程弘毅;;北大池剖面記錄的毛烏素沙地歷史沙漠化研究[A];地理學(xué)核心問題與主線——中國地理學(xué)會2011年學(xué)術(shù)年會暨中國科學(xué)院新疆生態(tài)與地理研究所建所五十年慶典論文摘要集[C];2011年
相關(guān)重要報紙文章 前4條
1 林勇 鄭怡;螺旋藻養(yǎng)殖技術(shù)（二）[N];福建科技報;2005年
2 林蔚學(xué) 鐘春云;大池工商所加強訂單農(nóng)業(yè)監(jiān)管[N];閩西日報;2005年
3 本報記者 殷磊邋通訊員 鄭怡婷;東部原水輸送工程讓廈門“左右逢源”[N];廈門日報;2007年
4 記者 余萬芳邋實習(xí)生 丁煌瓊;東大池社區(qū)豎起首塊公德牌[N];紹興日報;2007年
相關(guān)碩士學(xué)位論文 前3條
1 王治丹;基于最大池圖匹配的形變目標(biāo)跟蹤方法[D];合肥工業(yè)大學(xué);2016年
2 李衍鑫;枕大池注射示蹤劑觀察腦脊液淋巴回流及其臨床意義研究[D];吉林大學(xué);2010年
3 鐘嚴(yán)艷;UCMSCs經(jīng)枕大池移植治療發(fā)育鼠HIE后遺癥實驗研究[D];華中科技大學(xué);2011年
，本文編號：2139816
本文鏈接：http://sikaile.net/kejilunwen/ruanjiangongchenglunwen/2139816.html
最近更新
教材專著
