【摘要】：深度攝像頭的出現(xiàn)和日益完善,獲取物體的三維信息變得方便和快速,點(diǎn)云作為物體三維信息的一種重要表達(dá)形式,以點(diǎn)云為背景的計(jì)算機(jī)視覺(jué)近年來(lái)得到了發(fā)展。在許許多多的領(lǐng)域,三維視覺(jué)有著平面信息無(wú)法取代的作用,將機(jī)器視覺(jué)的應(yīng)用拓展到了新的領(lǐng)域。本文研究了以噴漆機(jī)器人對(duì)汽車(chē)保險(xiǎn)杠的識(shí)別和位姿估計(jì)相關(guān)的問(wèn)題,噴漆機(jī)器人對(duì)三維信息的依賴決定了三維視覺(jué)的必要性,點(diǎn)云的運(yùn)用使得噴漆機(jī)器人自動(dòng)完成對(duì)零件的識(shí)別成為可能,對(duì)于計(jì)算機(jī)視覺(jué)這個(gè)領(lǐng)域的進(jìn)一步發(fā)展,也是有著重要的意義。論文提出了三維點(diǎn)云的識(shí)別和位姿估計(jì)方案,包括點(diǎn)云處理和分割、點(diǎn)云識(shí)別、位姿估計(jì)三部分。首先決定了獲取點(diǎn)云的設(shè)備選擇,選擇kinect作為機(jī)器人的視覺(jué)硬件,并且人工獲取了各個(gè)保險(xiǎn)杠的全視角的完成點(diǎn)云。針對(duì)實(shí)驗(yàn)各個(gè)階段得到的三維點(diǎn)云,利用直通濾波和統(tǒng)計(jì)離群點(diǎn)濾波方法除去了比較明顯的噪聲點(diǎn),然后通過(guò)進(jìn)一步的稀疏濾波手段,獲得了比較適合后序處理的點(diǎn)云密度,并且由于實(shí)驗(yàn)的需要,提出一種特征點(diǎn)區(qū)別濾波方法,對(duì)于Thrift特征點(diǎn)周?chē)狞c(diǎn)云保持比較高的密度,離特征點(diǎn)比較遠(yuǎn)的部分保留比較稀疏的點(diǎn)云,并設(shè)置了對(duì)比實(shí)驗(yàn)驗(yàn)證效果。對(duì)全視角點(diǎn)云進(jìn)行模擬單視角采集,并計(jì)算這些單視角點(diǎn)云的視點(diǎn)特征直方圖VFH(Viewpoint Feature Histogram)特征計(jì)算,利用這些數(shù)據(jù)訓(xùn)練主成分分析SVM(Support Vector Machine)分類器。在識(shí)別和位姿估計(jì)階段,對(duì)于濾波處理后的點(diǎn)云數(shù)據(jù),選擇基于最小歐式距離的聚類分割方法實(shí)現(xiàn)了對(duì)單視角情況下點(diǎn)云數(shù)據(jù)的分割,并對(duì)分割后的各個(gè)聚類進(jìn)行提取視點(diǎn)特征直方圖VFH,然后利用已經(jīng)訓(xùn)練好的SVM分類器對(duì)這些VFH特征進(jìn)行分類。利用建立kd-tree(kdemention)和BP(Back Propagation)神經(jīng)網(wǎng)絡(luò)識(shí)別兩種手段進(jìn)行了位姿估計(jì)并進(jìn)行對(duì)比。其中,識(shí)別和位姿估計(jì)部分還分別設(shè)置了應(yīng)用主成分分析PCA(Principal Component Analysis)降維和不降維兩種方式的對(duì)比試驗(yàn)。實(shí)驗(yàn)結(jié)果表明本論文設(shè)計(jì)的點(diǎn)云預(yù)處理、分割識(shí)別和位姿估計(jì)具有可行性,能夠更快速的完成識(shí)別和位姿估計(jì)功能,有較大的探究?jī)r(jià)值。
[Abstract]:With the appearance and improvement of depth camera, it becomes more and more convenient and fast to obtain 3D information of objects. As an important expression of 3D information of objects, point cloud has been developed in computer vision with point cloud as the background in recent years. In many fields, 3D vision plays an irreplaceable role in plane information, which extends the application of machine vision to new fields. This paper studies the problems related to the identification and pose estimation of automobile bumper by painting robot. The dependence of painting robot on 3D information determines the necessity of 3D vision. The application of the point cloud makes it possible for the painting robot to recognize the parts automatically, which is of great significance to the further development of the field of computer vision. In this paper, three dimensional point cloud recognition and pose estimation schemes are proposed, including point cloud processing and segmentation, point cloud recognition and pose estimation. First, the selection of the device to obtain the point cloud is determined, and the kinect is chosen as the vision hardware of the robot, and the complete point cloud of the full angle of view of each bumper is obtained manually. For the 3D point cloud obtained in each stage of the experiment, the obvious noise points are removed by means of direct pass filtering and statistical outlier filtering, and the point cloud density suitable for post-sequence processing is obtained by further sparse filtering. Because of the need of experiment, a method of distinguishing feature points is proposed, which keeps a high density for the point cloud around the Thrift feature point, and keeps the sparse point cloud in the far part from the feature point, and sets up a comparative experiment to verify the effect. The full view point cloud is simulated and the view feature histogram (VFH (Viewpoint Feature Histogram) of these single view point clouds is calculated. The principal component analysis (SVM (Support Vector Machine) classifier is trained by these data. In the phase of recognition and pose estimation, the minimum Euclidean distance based clustering segmentation method is used to segment the point cloud data with single view angle. The view feature histogram (VFH,) is extracted from each clustering, and then the trained SVM classifier is used to classify these VFH features. Kd-tree (kdemention) and BP (Back Propagation) neural network recognition are used to estimate the position and pose. In the part of recognition and pose estimation, a comparative experiment of using principal component analysis (PCA) to reduce PCA (Principal Component Analysis) and not to reduce dimension is also carried out. The experimental results show that the point cloud preprocessing, segmentation recognition and pose estimation designed in this paper are feasible, and the functions of recognition and pose estimation can be completed more quickly.
【學(xué)位授予單位】：哈爾濱工業(yè)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;TP242

【參考文獻(xiàn)】

相關(guān)期刊論文 前10條

1 蘇本躍;馬金宇;彭玉升;盛敏;;基于K-means聚類的RGBD點(diǎn)云去噪和精簡(jiǎn)算法[J];系統(tǒng)仿真學(xué)報(bào);2016年10期

2 袁華;龐建鏗;莫建文;;基于體素化網(wǎng)格下采樣的點(diǎn)云簡(jiǎn)化算法研究[J];電視技術(shù);2015年17期

3 劉輝;王伯雄;任懷藝;李鵬程;;ICP算法在雙目結(jié)構(gòu)光系統(tǒng)點(diǎn)云匹配中的應(yīng)用[J];清華大學(xué)學(xué)報(bào)(自然科學(xué)版);2012年07期

4 譚志國(guó);魯敏;胡延平;郭裕蘭;莊釗文;;基于點(diǎn)云-模型匹配的激光雷達(dá)目標(biāo)識(shí)別[J];計(jì)算機(jī)工程與科學(xué);2012年04期

5 方旭;;基于BP神經(jīng)網(wǎng)絡(luò)人臉識(shí)別方法的研究與改進(jìn)[J];電腦知識(shí)與技術(shù);2011年04期

6 魏永超;劉長(zhǎng)華;杜冬;;基于曲面分割的三維點(diǎn)云物體識(shí)別[J];光子學(xué)報(bào);2010年12期

7 宇雪垠;曹拓荒;陳本盛;;基于特征臉的人臉識(shí)別及實(shí)現(xiàn)[J];河北工業(yè)科技;2009年05期

8 孫亞;;基于粒子群BP神經(jīng)網(wǎng)絡(luò)人臉識(shí)別算法[J];計(jì)算機(jī)仿真;2008年08期

9 梁新合;宋志真;;改進(jìn)的點(diǎn)云精確匹配技術(shù)[J];裝備制造技術(shù);2008年03期

10 白裔峰;肖建;于龍;黃景春;;基于結(jié)構(gòu)風(fēng)險(xiǎn)最小化的加權(quán)偏最小二乘法[J];計(jì)算機(jī)應(yīng)用;2007年04期

相關(guān)博士學(xué)位論文 前1條

1 蘇宏濤;基于統(tǒng)計(jì)特征的人臉識(shí)別技術(shù)研究[D];西北工業(yè)大學(xué);2004年

相關(guān)碩士學(xué)位論文 前6條

1 張楠;鐵路場(chǎng)景下三維點(diǎn)云識(shí)別與分類算法研究[D];北京交通大學(xué);2016年

2 常江;基于特征匹配的三維點(diǎn)云配準(zhǔn)算法研究[D];中北大學(xué);2015年

3 林志強(qiáng);面向智能服務(wù)機(jī)器人的物體感知研究[D];中國(guó)科學(xué)技術(shù)大學(xué);2014年

4 趙春雷;粗糙空間上結(jié)構(gòu)風(fēng)險(xiǎn)最小化原則[D];河北大學(xué);2011年

5 戴永前;基于二維激光雷達(dá)的移動(dòng)機(jī)器人三維環(huán)境的識(shí)別[D];南京理工大學(xué);2007年

6 孫宇;基于激光雷達(dá)的機(jī)器人三維地形構(gòu)建和草叢中障礙物檢測(cè)[D];浙江大學(xué);2007年

，

本文編號(hào)：2250025