發(fā)布時(shí)間：2018-09-12 06:59

【摘要】：洗車(chē)行為識(shí)別是復(fù)雜場(chǎng)景下人體行為識(shí)別的一個(gè)分支,目前簡(jiǎn)單場(chǎng)景下的人體簡(jiǎn)單動(dòng)作的識(shí)別已基本得到解決,而復(fù)雜場(chǎng)景下的行為識(shí)別仍面臨很多困難。洗車(chē)行中攝像頭的特殊角度拍攝人體動(dòng)作輪廓不分明以及汽車(chē)、工人頻繁移動(dòng)造成的“鬼影區(qū)域”都使得對(duì)洗車(chē)行為的識(shí)別更加困難。當(dāng)前傳統(tǒng)的行為識(shí)別算法并不能適應(yīng)洗車(chē)行這種特殊環(huán)境,針對(duì)洗車(chē)行為識(shí)別,本文提出了一種基于多核多示例的學(xué)習(xí)算法,以提高洗車(chē)行環(huán)境下洗車(chē)工人行為識(shí)別的準(zhǔn)確率。本文采用了改進(jìn)的ViBe背景差分法對(duì)運(yùn)動(dòng)目標(biāo)實(shí)時(shí)檢測(cè)來(lái)解決消除“鬼影區(qū)域”問(wèn)題,采用HOG-LBP特征提取算法來(lái)應(yīng)對(duì)人體動(dòng)作輪廓不分明的問(wèn)題。識(shí)別算法采用多核多示例學(xué)習(xí)算法,該算法將多核支持向量機(jī)與多示例學(xué)習(xí)算法有機(jī)結(jié)合,能夠有效的處理提取到的HOG-LBP融合特征,同時(shí)也提高了識(shí)別算法的學(xué)習(xí)能力,進(jìn)一步提高洗車(chē)行為識(shí)別的準(zhǔn)確率。實(shí)驗(yàn)證明,多核多示例學(xué)習(xí)算法在實(shí)驗(yàn)構(gòu)建的洗車(chē)行為數(shù)據(jù)集上,與傳統(tǒng)的行為識(shí)別算法相比,其識(shí)別率有所提高。本文中的算法是針對(duì)洗車(chē)行環(huán)境提出的,也同樣適用于類(lèi)似洗車(chē)行環(huán)境的復(fù)雜場(chǎng)景下的行為識(shí)別問(wèn)題。
[Abstract]:Car washing behavior recognition is a branch of human behavior recognition in complex scenes. At present, the recognition of simple human actions in simple scenes has been basically solved, but the behavior recognition in complex scenes still faces many difficulties. The special angle of the camera in the car washing shop makes it more difficult to recognize the car washing behavior because of the unclear outline of human body movement and the "ghost area" caused by the frequent movement of the workers. At present, the traditional behavior recognition algorithm can not adapt to the special environment of car washing line. In view of car washing line recognition, this paper proposes a learning algorithm based on multi-core and multi-example to improve the accuracy of car washing workers' behavior recognition in car washing environment. In this paper, the improved ViBe background differential method is used to detect moving objects in real time to solve the problem of eliminating the "ghost region", and the HOG-LBP feature extraction algorithm is used to deal with the problem of unclear human action contour. The recognition algorithm adopts multi-core and multi-example learning algorithm, which combines multi-core support vector machine with multi-example learning algorithm, which can deal with the extracted HOG-LBP fusion features effectively and improve the learning ability of the recognition algorithm. Further improve the car washing line for recognition accuracy. The experimental results show that the multi-core multi-example learning algorithm is more efficient than the traditional behavior recognition algorithm in the experimental data set. In this paper, the algorithm is proposed for car washing environment, and it is also applicable to the problem of behavior recognition in complex scenarios similar to car washing environment.
【學(xué)位授予單位】：哈爾濱工程大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2016
【分類(lèi)號(hào)】：U472.2;TP391.41

【參考文獻(xiàn)】

相關(guān)期刊論文 前10條

1 韓麗;黎琳;徐建國(guó);唐棣;;基于表面及空間特征的人體模型結(jié)構(gòu)分析[J];模式識(shí)別與人工智能;2015年03期

2 劉威;段成偉;遇冰;柴麗穎;袁淮;趙宏;;基于后驗(yàn)HOG特征的多姿態(tài)行人檢測(cè)[J];電子學(xué)報(bào);2015年02期

3 羅會(huì)蘭;單順勇;孔繁勝;;基于集成多示例學(xué)習(xí)的Mean Shift跟蹤算法[J];計(jì)算機(jī)輔助設(shè)計(jì)與圖形學(xué)學(xué)報(bào);2015年02期

4 李根;李文輝;;主方向旋轉(zhuǎn)LBP特征的平面旋轉(zhuǎn)人臉檢測(cè)[J];電子學(xué)報(bào);2015年01期

5 李武;胡冰;王明偉;;基于主成分分析和支持向量機(jī)的太赫茲光譜冰片鑒別[J];光譜學(xué)與光譜分析;2014年12期

6 田仙仙;鮑泓;徐成;;一種改進(jìn)HOG特征的行人檢測(cè)算法[J];計(jì)算機(jī)科學(xué);2014年09期

7 梁淑芬;劉銀華;李立琛;;基于LBP和深度學(xué)習(xí)的非限制條件下人臉識(shí)別算法[J];通信學(xué)報(bào);2014年06期

8 張靜;桑紅石;;基于初始尺度變換的SIFT匹配算法[J];紅外與毫米波學(xué)報(bào);2014年02期

9 瞿中;張亢;喬高元;;MB-LBP特征提取和粒子濾波相結(jié)合的運(yùn)動(dòng)目標(biāo)檢測(cè)與跟蹤算法研究[J];計(jì)算機(jī)科學(xué);2013年12期

10 徐少平;劉小平;李春泉;胡凌燕;楊曉輝;;基于LBP值對(duì)空間統(tǒng)計(jì)特征的紋理描述符[J];模式識(shí)別與人工智能;2013年08期

相關(guān)會(huì)議論文 前1條

1 王彪;王成儒;王芬芬;;一種改進(jìn)的運(yùn)動(dòng)目標(biāo)檢測(cè)算法[A];計(jì)算機(jī)技術(shù)與應(yīng)用進(jìn)展·2007——全國(guó)第18屆計(jì)算機(jī)技術(shù)與應(yīng)用（CACIS）學(xué)術(shù)會(huì)議論文集[C];2007年

相關(guān)碩士學(xué)位論文 前1條

1 于紀(jì)征;視頻中運(yùn)動(dòng)目標(biāo)的輪廓跟蹤與識(shí)別[D];上海交通大學(xué);2008年

，

本文編號(hào)：2238255