基于Camshift算法的運(yùn)動(dòng)目標(biāo)跟蹤研究

我的畢業(yè)論文題目是基于CamShift算法的運(yùn)動(dòng)目標(biāo)跟蹤，主要是對(duì)人體進(jìn)行跟蹤，精度可以跟蹤鉛筆、橡皮等小物件。包含了全套畢設(shè)成果包括文獻(xiàn)綜述、開(kāi)題報(bào)告、外文翻譯、畢業(yè)論文、附件清單、答辯ppt和程序代碼（使用VC+MFC編程，其中利用了opencv庫(kù)）。如果有不懂的地方，可以詢(xún)問(wèn)程序問(wèn)題。

摘  要

　　　目前，運(yùn)動(dòng)目標(biāo)跟蹤技術(shù)已經(jīng)成為了計(jì)算機(jī)領(lǐng)域與人工智能領(lǐng)域中最熱門(mén)的研究課題之一，在軍事和民用領(lǐng)域中有著廣泛的應(yīng)用，具有極其重要的研究意義和研究?jī)r(jià)值，而Camshift運(yùn)動(dòng)目標(biāo)跟蹤算法則是其中一種非常重要的算法。
　　　本文闡述了目標(biāo)跟蹤技術(shù)的基本理論，并著重介紹了Camshift運(yùn)動(dòng)目標(biāo)跟蹤算法的基本原理和方法，同時(shí)通過(guò)MFC和Opencv實(shí)現(xiàn)了運(yùn)動(dòng)目標(biāo)跟蹤，通過(guò)實(shí)驗(yàn)所得對(duì)比圖測(cè)試和比較了Meanshift算法和Camshift算法的效果，從而得出Camshift算法的優(yōu)點(diǎn)和缺陷，進(jìn)而對(duì)Camshift算法的應(yīng)用和改進(jìn)進(jìn)行了展望。
　　　實(shí)驗(yàn)結(jié)果表明，Camshift算法具有計(jì)算效率高，魯棒性好，實(shí)時(shí)性強(qiáng)等特點(diǎn)，可以有效的解決目標(biāo)局部被遮擋、變形等問(wèn)題，跟蹤效果很好，但同時(shí)Camshift算法在目標(biāo)運(yùn)動(dòng)速度過(guò)快、光照強(qiáng)烈等條件下容易丟失目標(biāo)。

關(guān)鍵詞：Camshift；運(yùn)動(dòng)目標(biāo)跟蹤；Opencv；MFC

ABSTRACT

　　　At present, the technology of Moving Objects Tracking has become one of the most popular research topics in the field of Computer Science and Artificial Intelligence. It has been widely used in the field of Military and Civil Application, and has an extremely important research significance and value, while the Moving Objects Tracking Using by Camshift Algorithm is one of the important Moving Objects Tracking Algorithms.
　　　This paper expounds the basic theory of Objects Tracking technology, and introduces the basic principle and method of Moving Objects Tracking Using by Camshift Algorithm emphatically. At the same time, it realized Objects Tracking by MFC and Opencv, testing and compared the effect of Meanshift Algorithm and Camshift Algorithm by the contrast diagram obtained through the experiment, thus gets the advantages and disadvantages of Camshift Algorithm and discussed the application improvement of Camshift Algorithm.
　　Experimental results show that Camshift Algorithm has high computational efficiency, good robustness and real-time. It has good tracking results, and can solve the problem of target obscured effectively. However, when the speed of Moving Objects are too fast or light is strong, it will be easy to lose objects.
　　　

Keywords: Camshift；Moving Objects Tracking；Opencv；MFC

目  錄
摘  要
Abstract
第1章 緒論 1
1.1 選題意義和論文任務(wù) 1
1.2 國(guó)內(nèi)外研究進(jìn)展 1
1.3 論文結(jié)構(gòu) 3
第2章 運(yùn)動(dòng)目標(biāo)跟蹤算法原理 4
2.1引言 4
2.2基于相關(guān)的目標(biāo)跟蹤算法 5
2.3基于光流的目標(biāo)跟蹤算法 5
2.4基于特征的目標(biāo)跟蹤算法 6
2.4.1 Meanshift算法 6
2.4.2 Camshift算法 9
2.5基于假設(shè)條件的目標(biāo)跟蹤算法 10
第3章 基于Camshift算法的運(yùn)動(dòng)目標(biāo)跟蹤研究 10
3.1 反投影計(jì)算 10
3.2 Meanshift算法 15
3.3 Camshift算法 15
第4章 基于Camshift算法的運(yùn)動(dòng)目標(biāo)跟蹤系統(tǒng)軟件設(shè)計(jì)與實(shí)現(xiàn) 16
4.1 界面設(shè)計(jì) 16
4.2 效果測(cè)試 17
4.2.1程序基礎(chǔ)功能測(cè)試 17
4.2.2 Meanshift算法效果測(cè)試 21
4.2.3 Camshift算法效果測(cè)試 24
4.2.4 Meanshift算法與Camshift算法性能對(duì)比分析 26
第5章 總結(jié)與展望 27
5.1總結(jié) 27
5.2展望 27
致謝 29
參考文獻(xiàn) 30