本文選題：壓縮感知 + 目標(biāo)檢測(cè)��； 參考：《北京郵電大學(xué)》2016年碩士論文

【摘要】：多目標(biāo)實(shí)時(shí)跟蹤是計(jì)算機(jī)視覺(jué)領(lǐng)域的研究熱點(diǎn)之一,在智能交通、智能監(jiān)控等多個(gè)領(lǐng)域有著廣泛的應(yīng)用,然而設(shè)計(jì)一個(gè)魯棒性高的多目標(biāo)實(shí)時(shí)跟蹤系統(tǒng),無(wú)論是對(duì)科學(xué)研究還是工程實(shí)踐都極具挑戰(zhàn)性。壓縮感知算法通過(guò)對(duì)信號(hào)采樣壓縮,能夠大大降低信號(hào)的復(fù)雜度。將壓縮感知理論與多目標(biāo)實(shí)時(shí)跟蹤相結(jié)合能夠提升系統(tǒng)跟蹤的穩(wěn)定性和實(shí)時(shí)性。本文基于壓縮感知理論,設(shè)計(jì)了一套魯棒性高的多目標(biāo)實(shí)時(shí)跟蹤系統(tǒng)。主要研究?jī)?nèi)容如下:(1)研究并實(shí)現(xiàn)了基于Haar特征的AdaBoost目標(biāo)檢測(cè)器。提取視頻圖像的Haar特征,并基于大量多角度人頭正負(fù)樣本圖像訓(xùn)練AdaBoost級(jí)聯(lián)分類器,實(shí)現(xiàn)多目標(biāo)檢測(cè)。(2)設(shè)計(jì)并改進(jìn)了基于壓縮感知的樸素貝葉斯目標(biāo)跟蹤器�；趬嚎s感知理論,對(duì)目標(biāo)特征采樣獲得壓縮感知特征,并構(gòu)建分特.征權(quán)重的樸素貝葉斯在線學(xué)習(xí)分類器,實(shí)現(xiàn)多目標(biāo)實(shí)時(shí)跟蹤。(3)開(kāi)發(fā)了一套多目標(biāo)實(shí)時(shí)跟蹤系統(tǒng),在多種實(shí)驗(yàn)場(chǎng)景下與MHT算和CT算法進(jìn)行檢測(cè)跟蹤實(shí)驗(yàn)比較,證明本系統(tǒng)在實(shí)時(shí)性、準(zhǔn)確性和穩(wěn)定性上都有較好的表現(xiàn)。本文將壓縮感知算法應(yīng)用到目標(biāo)跟蹤系統(tǒng)中,實(shí)現(xiàn)了目標(biāo)跟蹤系統(tǒng)穩(wěn)定性和實(shí)時(shí)性的兼?zhèn)涞亩嗄繕?biāo)實(shí)時(shí)跟蹤系統(tǒng),具有很高的工程研究?jī)r(jià)值和社會(huì)應(yīng)用價(jià)值。
[Abstract]:Multi-target real-time tracking is one of the hotspots in the field of computer vision. It has been widely used in many fields, such as intelligent transportation, intelligent monitoring and so on. However, a robust multi-target real-time tracking system is designed. Both scientific research and engineering practice are extremely challenging. The compression sensing algorithm can greatly reduce the complexity of the signal by sampling and compressing the signal. Combining compression sensing theory with multi-target real-time tracking can improve the stability and real-time of system tracking. Based on the theory of compressed sensing, a robust multi-target real-time tracking system is designed in this paper. The main contents are as follows: (1) AdaBoost target detector based on Haar feature is studied and implemented. The Haar feature of video image is extracted and the AdaBoost cascade classifier is trained based on a large number of multi-angle head positive and negative samples. (2) A naive Bayesian target tracker based on compressed sensing is designed and improved. Based on the theory of compressed perception, the compressed perceptual features are obtained by sampling the target features, and the sub-features are constructed. Naive Bayesian online learning classifier with eigenweight is used to realize multi-target real-time tracking. (3) A multi-target real-time tracking system is developed and compared with MHT algorithm and CT algorithm in various experimental scenarios. It is proved that the system has good performance in real time, accuracy and stability. In this paper, the compressed sensing algorithm is applied to the target tracking system, and the multi-target real-time tracking system is realized, which is both stable and real-time. It has high engineering research value and social application value.
【學(xué)位授予單位】：北京郵電大學(xué)
【學(xué)位級(jí)別】：碩士
【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 王松林;項(xiàng)欣光;;基于壓縮感知的多特征加權(quán)目標(biāo)跟蹤算法[J];計(jì)算機(jī)應(yīng)用研究;2014年03期

2 Lizuo Jin;Tirui Wu;Feng Liu;Gang Zeng;;Hierarchical Template Matching for Robust Visual Tracking with Severe Occlusions[J];ZTE Communications;2012年04期

3 Simon X.Yang;;Fast-moving target tracking based on mean shift and frame-difference methods[J];Journal of Systems Engineering and Electronics;2011年04期

，

本文編號(hào)：2103017