(44.192.70.216) 您好!臺灣時間:2021/05/09 18:24
字體大小: 字級放大   字級縮小   預設字形  
回查詢結果

詳目顯示:::

我願授權國圖
: 
twitterline
研究生:彭柏嘉
研究生(外文):Bo-Jia Peng
論文名稱:一種基於機器學習之混合粒子濾波法於目標追蹤機器人之應用
論文名稱(外文):A Machine Learning-based Hybrid Particle Filter for Object Tracking Robot Applications
指導教授:周佑誠
指導教授(外文):Yu-Cheng Chou
學位類別:碩士
校院名稱:中原大學
系所名稱:機械工程研究所
學門:工程學門
學類:機械工程學類
論文種類:學術論文
論文出版年:2014
畢業學年度:102
語文別:中文
論文頁數:71
中文關鍵詞:機器視覺運動控制移動機器人機器學習目標追蹤倒傳遞類神經網路AdaBoost粒子濾波
外文關鍵詞:Machine Vision-based Motion ControlMobile RobotObject TrackingMachine LearningBack Propagation Neural NetworkAdaBoostParticle Filter
相關次數:
  • 被引用被引用:0
  • 點閱點閱:213
  • 評分評分:系統版面圖檔系統版面圖檔系統版面圖檔系統版面圖檔系統版面圖檔
  • 下載下載:0
  • 收藏至我的研究室書目清單書目收藏:0
在機器視覺的領域目標偵測或追蹤一直是很熱門的研究問題,已經有許多追蹤框架被提出且被應用於各個領域。粒子濾波已經被證實是一種有效的追蹤框架,尤其當目標運動行為具有非線性、非高斯特性時能達到很好的追蹤效果,因此非常適合實際應用。
本論文提出一種基於機器學習混合粒子濾波法追蹤系統,藉由結合AdaBoost分類器與傳統粒子濾波法,以改善傳統粒子濾波法之目標追蹤效率。首先利用AdaBoost分類器可以快速判斷前景(目標物)與背景的特性,偵測目標物在影像中的位置並周期性地更新目標物的參考色彩直方圖供粒子濾波器使用,接著由粒子濾波器對目標物作預測、量測的動作,找出最有可能是目標物的位置,最後由倒傳遞類神經網路(Back Propagation Neural Network,BPN)所訓練的垂直深度與轉換比例關係將目標物的影像資訊轉換為真實的座標資訊,以達到追蹤之目的。
本論文進行了在不同粒子數與不同光照環境下的L型追蹤,以及目標物消失後再出現之不同追蹤實驗,並與傳統粒子濾波法比較。由實驗結果顯示,本論文所提出之方法在以上實驗能達到良好的追蹤效果,且效果優於傳統粒子濾波法。

This thesis presents a machine learning-based hybrid particle filter for object tracking robot applications. The presented hybrid particle filter integrates an AdaBoost classifier with a traditional particle filter, in order to improve the traditional particle filter’s object tracking performance under different environmental configurations.
In the presented object tracking method for monocular vision-based follower robots, a well-trained AdaBoost classifier periodically detects the object’s location in an image and generates a new reference color histogram for the object. The new reference color histogram is then used by a traditional particle filter to perform the prediction, measurement, and sorting operations, in order to produce the moving object’s most possible location in an image at the current time-step. The obtained location is then passed to two well-trained back propagation neural networks to generate the moving object’s polar coordinate, including the distance and angle, relative to the follower robot. Eventually, the obtained polar coordinate is used to create speed and rotation commands, which make the follower robot move towards the object in order to keep the moving object in the camera’s central field of view.
In this thesis, different experiments, including the L-shape tracking under different number of particles and different lighting conditions, and the tracking of an object that temporally disappears from the camera’s field of view, are conducted using both the presented hybrid particle filter and traditional particle filter. Experimental results show that the presented hybrid particle filter performs well and significantly better than the traditional particle filter in the above object tracking scenarios.

目錄
摘要 I
Abstract II
致謝 IV
目錄 V
圖目錄 VII
表目錄 X
第一章 緒論 1
1.1前言 1
1.2文獻回顧 1
1.3研究方法 2
1.4論文架構 3
第二章 研究理論 4
2.1 機器學習 4
2.2 AdaBoost演算法 4
2.3 Haar-like特徵 6
2.3.1積分影像 8
2.4粒子濾波 10
2.5類神經網路 11
2.5.1倒傳遞類神經網路 13
第三章 基於機器學習之混合粒子濾波法追蹤系統 18
3.1混合粒子濾波追蹤系統流程 18
3.2 AdaBoost分類器之訓練與目標物辨識 19
3.3 HSV色彩空間 20
3.4 目標物追蹤 21
3.5 目標物實際距離計算 24
3.5.1垂直深度映射 24
3.5.2水平距離映射 27
3.6追隨機器人之運動控制 28
第四章 移動機器人軟硬體系統 33
4.1系統架構 33
4.2感測系統 34
4.3移動平台 35
4.4控制系統 36
4.5軟體環境 37
第五章 目標追蹤實驗與討論 38
5.1 L型追蹤實驗 38
5.2不同粒子數實驗 44
5.3不同光照環境實驗 49
5.4目標物消失再出現實驗 52
5.5實驗結果討論 55
第六章 結論與未來展望 57
6.1結論 57
6.2未來展望 58
參考文獻 59

圖目錄
圖2.1 機器學習示意圖 4
圖2.2 AdaBoost演算法 6
圖2.3 矩形特徵與人臉特徵的關係 6
圖2.4 Haar-like特徵與AmigoBot機器人特徵的關係 8
圖2.5 積分影像計算範例 9
圖2.6 計算任意矩形的積分影像值 9
圖2.7 以粒子濾波進行基於影像的目標物追蹤流程圖 11
圖2.8 倒傳遞類神經網路架構圖 14
圖2.9 倒傳遞類神經網路計算流程圖 15
圖3.1 混合粒子濾波追蹤系統流程圖 19
圖3.2 正樣本(a) ~ (b) 19
圖3.3 負樣本(a) ~ (b) 19
圖3.4 AdaBoost分類器之訓練與目標物辨識流程圖 20
圖3.5 分類器目標辨識結果(a) ~ (b) 20
圖3.6 HSV色彩空間 21
圖3.7 結合AdaBoost分類器與粒子濾波進行基於影像的目標物追蹤流程圖 23
圖3.8 實際距離、垂直深度和水平距離的幾何關係圖 24
圖3.9 在影像中判斷物體的像素高與水平像素寬 26
圖3.10 倒傳遞類神經網路垂直深度映射數據訓練 26
圖3.11 倒傳遞類神經網路轉換比例數據訓練 28
圖3.12 角度θ與水平距離、垂直深度的幾何關係圖 29
圖3.13 控制追隨機器人速度的判斷基準 30
圖3.14 基於機器學習之混合粒子濾波法追蹤系統流程圖 31
圖4.1 機器人硬體系統架構 33
圖4.2 追隨機器人 34
圖4.3 AXIS 207MW Network Cameras 34
圖4.4 AmigoBot 36
圖4.5 VIA ARTIGO A1100 36
圖5.1 L型追蹤示意圖 38
圖5.2 求取追隨機器人理想行進路徑之方法示意圖 39
圖5.3 追隨機器人的理想行進路徑 39
圖5.4 使用粒子濾波於L型追蹤影像流程 40
圖5.5 使用粒子濾波時追隨機器人與目標路徑圖(粒子數100) 41
圖5.6 結合粒子濾波與AdaBoost分類器於L型追蹤影像流程 42
圖5.7 結合粒子濾波與AdaBoost分類器追隨機器人與目標路徑圖(粒子數100) 43
圖5.8 使用粒子濾波於不同粒子數的L型實驗結果 46
圖5.9 結合粒子濾波與AdaBoost分類器於不同粒子數的L型實驗結果 48
圖5.10 實驗環境光源配置圖 49
圖5.11 關第4盞燈時L型追蹤實驗結果 49
圖5.12 關第3盞燈時L型追蹤實驗結果 50
圖5.13 關第3,4盞燈時L型追蹤實驗結果 50
圖5.14 關第1,3,4盞燈時L型追蹤實驗結果 51
圖5.15 使用粒子濾波之影像序列追蹤流程 53
圖5.16 結合粒子濾波與AdaBoost分類器之影像序列追蹤流程 54

表目錄
表2.1 常用的Haar-like特徵 7
表3.1 不同垂直深度下物體的像素高 25
表3.2 正規化後之不同垂直深度下物體的像素高 25
表3.3 不同垂直深度下的轉換比例 27
表3.4 正規化後之不同垂直深度下的轉換比例 27
表4.1 AXIS 207MW攝影機規格表 35
表4.2 VIA ARTIGO A1100嵌入式電腦規格表 37

1.Ng, G.W. and G. Ng, Intelligent systems: fusion, tracking and control. 2003: Research Studies Press.
2.Shi, J. and C. Tomasi. Good features to track. in Computer Vision and Pattern Recognition, 1994. Proceedings CVPR'94., 1994 IEEE Computer Society Conference on. 1994: IEEE.
3.Comaniciu, D., V. Ramesh, and P. Meer, Kernel-based object tracking. Pattern Analysis and Machine Intelligence, IEEE Transactions on, 2003. 25(5): p. 564-577.
4.Collins, R.T., Y. Liu, and M. Leordeanu, Online selection of discriminative tracking features. Pattern Analysis and Machine Intelligence, IEEE Transactions on, 2005. 27(10): p. 1631-1643.
5.Arulampalam, M.S., et al., A tutorial on particle filters for online nonlinear/non-Gaussian Bayesian tracking. Signal Processing, IEEE Transactions on, 2002. 50(2): p. 174-188.
6.Foresti, G.L., et al. Face detection for visual surveillance. in Image Analysis and Processing, 2003. Proceedings. 12th International Conference on. 2003: IEEE.
7.Lee, J.M., et al., Localization of a mobile robot using the image of a moving object. Industrial Electronics, IEEE Transactions on, 2003. 50(3): p. 612-619.
8.Shan, C., et al. Real time hand tracking by combining particle filtering and mean shift. in Automatic Face and Gesture Recognition, 2004. Proceedings. Sixth IEEE International Conference on. 2004: IEEE.
9.Kwolek, B., Stereovision-based head tracking using color and ellipse fitting in a particle filter, in Computer Vision-ECCV 2004. 2004, Springer. p. 192-204.
10.Grabner, H., M. Grabner, and H. Bischof. Real-Time Tracking via On-line Boosting. in BMVC. 2006.
11.Grabner, H. and H. Bischof. On-line boosting and vision. in Computer Vision and Pattern Recognition, 2006 IEEE Computer Society Conference on. 2006: IEEE.
12.Okuma, K., et al., A boosted particle filter: Multitarget detection and tracking, in Computer Vision-ECCV 2004. 2004, Springer. p. 28-39.
13.Zhao, F., G. Liu, and X. Wang. An efficient AdaBoost tracking algorithm based on the particle framework. in Signal Processing, Communications and Computing (ICSPCC), 2011 IEEE International Conference on. 2011: IEEE.
14.Iwahori, Y., et al. Efficient tracking with Adaboost and particle filter under complicated background. in Knowledge-Based Intelligent Information and Engineering Systems. 2008: Springer.
15.Fahn, C.-S., M.-J. Kuo, and K.-Y. Wang, Real-time face tracking and recognition based on particle filtering and AdaBoosting techniques, in Human-Computer Interaction. Novel Interaction Methods and Techniques. 2009, Springer. p. 198-207.
16.Lei, Y., X. Ding, and S. Wang. Adaboost tracker embedded in adaptive particle filtering. in Pattern Recognition, 2006. ICPR 2006. 18th International Conference on. 2006: IEEE.
17.LI, T., et al., Robust Object Tracking by Combining Boosting Learning and Particle Filter. Journal of Computational Information Systems, 2012. 8(22): p. 9593-9601.
18.Schapire, R.E., The strength of weak learnability. Machine learning, 1990. 5(2): p. 197-227.
19.Freund, Y. and R.E. Schapire. A desicion-theoretic generalization of on-line learning and an application to boosting. in Computational learning theory. 1995: Springer.
20.Viola, P. and M. Jones. Rapid object detection using a boosted cascade of simple features. in Computer Vision and Pattern Recognition, 2001. CVPR 2001. Proceedings of the 2001 IEEE Computer Society Conference on. 2001: IEEE.
21.Viola, P. and M.J. Jones, Robust real-time face detection. International journal of computer vision, 2004. 57(2): p. 137-154.
22.Rumelhart, D.E., G.E. Hinton, and R.J. Williams, Learning internal representations by error propagation. 1985, DTIC Document.
23.Specht, D.F., Probabilistic neural networks. Neural networks, 1990. 3(1): p. 109-118.
24.Hecht-Nielsen, R., Counterpropagation networks. Applied optics, 1987. 26(23): p. 4979-4983.
25.Kohonen, T., Self-organized formation of topologically correct feature maps. Biological cybernetics, 1982. 43(1): p. 59-69.
26.Carpenter, G.A. and S. Grossberg, Adaptive resonance theory. 2010: Springer.
27.Hopfield, J.J., Neural networks and physical systems with emergent collective computational abilities. Proceedings of the national academy of sciences, 1982. 79(8): p. 2554-2558.
28.Kosko, B., Bidirectional associative memories. Systems, Man and Cybernetics, IEEE Transactions on, 1988. 18(1): p. 49-60.
29.AXIS 207MW Network Camera. Available from: http://www.axis.com/products/cam_207mw/.
30.AmogoBot Robot for Education and Research. Available from: http://www.mobilerobots.com/ResearchRobots/AmigoBot.aspx.
31.VIA ARTiGO A1100 PC Kit. Available from: http://www.via.com.tw/en/products/embedded/artigo/a1100/index.jsp.
32.VMware Workstation. Available from: http://www.vmware.com/tw/support/product-support/workstation.html.
33.Ubuntu for you. Available from: http://www.ubuntu.com/ubuntu.
34.OpenCV (Open Source Computer Vision). Available from: http://opencv.willowgarage.com/wiki/.
35.ARIA (Advanced Robot Interface for Applications). Available from: http://robots.mobilerobots.com/wiki/ARIA.

QRCODE
 
 
 
 
 
                                                                                                                                                                                                                                                                                                                                                                                                               
第一頁 上一頁 下一頁 最後一頁 top
系統版面圖檔 系統版面圖檔