臺灣博碩士論文加值系統

本論文基於電腦視覺技術提出了一套室內安全巡邏自動車系統，採用一台載有PTZ網路攝影機及具無線遙控功能的小型自動車作為實驗平台。本論文首先提出了一個易於使用的學習方法，供系統作學習巡邏環境之運用，其中包括：巡邏路線、地板顏色、監控物品，以及自動車相對於監控物品之位置。接著，本論文提出一個自動車安全巡邏的方法使自動車能進行安全監控以及障礙物自動閃避的工作。自動車會根據學習到的路徑作巡邏，並利用本論文所提出的簡化式SIFT(Scale Invariant Feature Transform)方法進行物品監控。該方法從含有監控物品的影像中抽取特徵點，並與學習所得版本進行比對，再藉由霍夫轉換找出兩版本間的仿射轉換，據以測定自動車相對於物品之位置，以及修正自動車航行路徑的偏移。此外，本論文也利用地板的代表顏色來偵測障礙物，讓自動車適應於花色地板的環境，並整合目標導向路徑追隨之策略，使車子能自動閃避障礙物。最後我們利用一實際的室內環境來測驗本論文所提出的系統，結果自動車能順利自動修正路徑以及監控物品，顯示出本論文所提方法的完整性以及可行性。

A vision-based vehicle system for security patrolling in indoor environments using an autonomous vehicle is proposed. A small vehicle with wireless control and a web camera which has the capabilities of panning, tilting, and zooming is used as a test bed. At first, an easy-to-use learning technique is proposed, which has the capability of extracting specific features, including navigation path, floor color, monitored object, and vehicle location with respect to monitored objects. Next, a security patrolling method by vehicle navigation with obstacle avoidance and security monitoring capabilities is proposed. The vehicle navigates according to the node data of the path map which is created in the learning phase and monitors concerned objects by a simplified scale-invariant feature transform (simplified-SIFT) algorithm proposed in this study. Accordingly, we can extract the features of each monitored object from acquired images and match them with the corresponding learned data by the Hough transform. Furthermore, a vehicle location estimation technique for path correction utilizing the monitored object matching result is proposed. In addition, techniques for obstacle avoidance are also proposed, which can be used to find the clusters of floor colors, detect obstacles in environments with various floor colors, and integrate a technique of goal-directed minimum path following to guide the vehicle to avoid obstacles. Good experimental results show the flexibility and feasibility of the proposed methods for the application of security patrolling in indoor environments.

ABSTRACT i
摘要 iii
ACKNOWLEDGEMENTS iv
CONTENTS v
LIST OF FIGURES viii
Chapter 1 Introduction 1
1.1 Motivation 1
1.2 Survey of Related Studies 3
1.3 Overview of Proposed Approach 5
1.4 Contributions 8
1.5 Thesis Organization 8
Chapter 2 System Configuration and Navigation Principles 10
2.1 Introduction 10
2.2 System Configuration 12
2.2.1 Hardware configuration 12
2.2.2 Software configuration 15
2.3 Learning Principle and Proposed Process 15
2.4 Vehicle Guidance Principle and Proposed Process 18
Chapter 3 Learning Strategies for Indoor Navigation by Manual Driving 20
3.1 Ideas of Proposed Techniques Used in Learning 20
3.1.1 Learning camera parameters 20
3.1.2 Environment features for learning 21
3.2 Camera Calibration 22
3.2.1 Calibration of location mapping 22
3.2.2 Calibration of intrinsic parameters 23
3.3 Learning of Specific Features 24
3.3.1 Learning of navigation paths composed of nodes 24
3.3.2 Learning of floor colors by k-means clustering 27
3.3.3 Learning of monitored objects by simplified SIFT 29
3.3.4 Learning of vehicle locations with respect to monitored objects 32
Chapter 4 Security Patrolling by Vehicle Navigation in Indoor Environments 36
4.1 Introduction to Proposed Ideas 36
4.2 Ideas of Navigation Process 37
4.2.1 Guidance by learned paths 37
4.2.2 Obstacle avoidance in various floor environments 39
4.2.3 Path correction by monitored object image matching 41
4.3 Purposes of Security Patrolling 44
4.3.1 Proposed technique for monitoring of objects 44
4.3.2 Types of monitored objects 46
4.4 Detailed Process for Security Patrolling by Vehicle Navigation 47
Chapter 5 Detection of Monitored Objects by 2D Object Image Matching 50
5.1 Introduction 50
5.2 Review of Method of Matching by Scale-Invariant Feature Transform (SIFT) 51
5.2.1 Scale-invariant keypoint localization 52
5.2.2 Feature descriptor generation 54
5.3 Proposed Simplified SIFT Features for Detection of Monitored Objects 56
5.3.1 Necessity of simplification of original concept 56
5.3.2 Detailed process of simplified SIFT feature generation 57
5.4 Proposed Matching Technique Using Simplified SIFT Features 62
5.4.1 Concept of proposed matching algorithm 62
5.4.2 Detailed algorithm 63
5.5 Experimental Results 65
Chapter 6 Vehicle Guidance by Location Estimation Based on 2D Object Image Matching Results 69
6.1 Introduction 69
6.2 Review of Vehicle Location Estimation Techniques 70
6.3 Vehicle Location Estimation by Object Image Matching Results 72
6.3.1 Coordinate systems 72
6.3.2 Idea of proposed method 73
6.3.3 Detailed process of location estimation 75
6.4 Path Correction by Vehicle Location Estimation Results 79
6.4.1 Direction angle of vehicle and coordinates of path nodes 79
6.4.2 Method of proposed path correction 81
6.5 Experimental Results 86
Chapter 7 Obstacle Avoidance in Various Floor Environments 90
7.1 Overview of Obstacle Avoidance Methods 90
7.2 Idea of Proposed Obstacle Avoidance Method 91
7.2.1 Coordinate system and direction angle of vehicle 91
7.2.2 Coordinate transformation 92
7.3 Obstacle Avoidance Techniques 93
7.3.1 Finding floor colors by k-means clustering 93
7.3.2 Obstacle detection by alert line scanning 96
7.3.3 Computation of goal-directed minimum path 99
7.4 Experimental Results 102
Chapter 8 Experimental Results and Discussions 105
8.1 Experimental Results 105
8.2 Discussions 111
Chapter 9 Conclusions and Suggestions for Future Works 113
9.1 Conclusions 113
9.2 Suggestions for Future Works 114
References 116

[1] G.N. DeSouza and A.C. Kak, “Vision for mobile robot navigation: A survey,” IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. 24, No 2, pp. 237-267, February 2002.
[2] C. C. Lai, “A study on automatic indoor navigation techniques for vision-based mini-vehicle with off-line environment learning capability,” M. S. Thesis, Department of Computer and Information Science, National Chiao Tung University, Taiwan, June, 2003.
[3] K. L. Chiang, “Security patrolling and danger condition monitoring in indoor environments by vision-based autonomous vehicle navigation,” M. S. Thesis, Institute of Computer Science and Engineering, National Chiao Tung University, Taiwan, June, 2006.
[4] Y. C. Chen and W. H. Tsai, “Vision-based autonomous vehicle navigation in complicated room environments with collision avoidance capability by simple learning and fuzzy guidance techniques,” Proceedings of 2004 Conference on Computer Vision, Graphics and Image Processing, Hualien, Taiwan, Republic of China, Aug. 2004.
[5] M. C. Chen and W. H. Tsai “Vision-based security patrolling in indoor environments using autonomous vehicles,” Proceedings of 2005 Conference on Computer Vision, Graphics and Image Processing, Taipei, Taiwan, Republic of China, Aug, 2005.
[6] C. Shmid and R. Mohr. “Local greyvalue invariants for image retrieval,” IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. 19, No. 5, pp. 530-534, 1997.
[7] K. Mikoljczyk and C. Schmid. “Indexing based on scale-invariant features,” in Proceedings of International Conference on Computer Vision, pp. 525-531,2001.
[8] D. G. Lowe, “Distinctive image features from scale-invariant keypoints,” International Journal of Computer Vision, Vol. 60, No. 2, pp. 91-110, 2004.
[9] I. Fukui, “TV image processing to determine the position of a robot vehicle,” Pattern Recognition, Vol. 14, pp. 101-109, 1981.
[10] M. J. Magee and J. K. Aggarwal, “Determining the position of a robot using a single calibration object,” Proceedings of IEEE Conference on Robotics, pp. 57-62, Atlanta, Georgia, May 1983.
[11] J. Huang, C. Zhao, Y. Ohtake, H. Li, and Q. Zhao, “Robot position identification using specially designed landmarks,” Proceedings of 2006 IEEE Conference on Instrumentation and Measurement Technology, April 2006.
[12] H. L. Chou and W. H. Tsai “A new approach to robot location by house corners,” Pattern Recognition, Vol. 19, pp. 439-451, 1986.
[13] K. L. Chiang and W. H. Tsai, “Vision-based autonomous vehicle guidance in indoor environments using odometer and house corner location information,” Proceedings of 2006 IEEE International Conference on Intelligent Information Hiding and Multimedia Signal Processing, pp. 415-418, Dec. 18 - 20, 2006.
[14] C. Shu, A. Brunton, and M. Fiala, “CAMcal: A program for camera calibration using checkerboard patterns,” Proceedings of 9th IEEE International Conference on Computer Vision (ICCV2003) demo session, CDROM proceedings, Nice, France, October, 2003.
[15] K. Mikolajczyk, and C. Schmid, “A performance evaluation of local descriptors,” Proceedings of International Conference on Computer Vision & Pattern Recognition, Vol. 2, pp. 257-263, June 2003.
[16] D. G. Lowe, “Local feature view clustering for 3D object recognition,” IEEE Conference on Computer Vision and Pattern Recognition, Kauai, Hawaii, pp. 682-688, December 2001.
[17] C. H. Ku and W. H. Tsai, “Obstacle avoidance in person following for vision-based autonomous land vehicle guidance using vehicle location estimation and quadratic pattern classifier,” IEEE Transactions on Industrial Electronics, Vol. 48, No. 1, pp. 205-215, 2001.
[18] E. Forgey, “Cluster analysis of multivariate data: Efficiency vs. interpretability of classification,” Biometrics, Vol. 21, pp. 768, 1965.
[19] J. MacQueen, “Some methods for classification and analysis of multivariate observations,” Proceedings of Fifth Berkeley Symposium Math. Statistics and Probability, Vol. 1, pp. 281-296, 1967.
[20] T. Kanungo, D. M. Mount, N. Netanyahu, C. Piatko, R. Silverman, and A. Y. Wu, “An efficient k-means clustering algorithm: Analysis and implementation,” IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. 24, No. 7, pp. 881-892, 2002.
[21] S. P. Lloyd, “Least Squares Quantization in PCM,” IEEE Transactions on Information Theory, Vol. 28, pp. 129-137, 1982.
[22] L. Lucchese , S. K. Mitra, “Unsupervised segmentation of color images based on k-means clustering in the chromaticity plane,” Proceedings of IEEE Workshop on Content-Based Access of Image and Video Libraries, pp.74, June 22-22, 1999.