视觉导航中环境建模的研究
|
摘要
自然环境中视觉导航的研究涉及到计算机视觉中的各个主要方面，是一个有
难度的综合性课题。视觉导航的基本任务是全局定位、道路跟踪和障碍物检测，
其中每一部分都要以环境建模为基础。本文系统深入地研究了环境建模问题，包
括：
1．提出了面向任务的多尺度和全覆盖视野的环境综合建模方法。将路标定位
的全景图象建模，道路识别的全方位图象建模和目标（障碍物）检测的双目注视图
象建模相结合，改进了依赖局部、单一信息的视觉导航方法。研究了面向任务的
环境建模方法中的传感器设计、视觉算法和模型表示问题。
2．提出了由非精确摄象机运动下的图象序列建立3D环境全景模型的两步
法，即基于运动滤波的图象稳定和基于时- 空－频域遮挡模型的全景外极面图象分
析。本文推广了全景图象方法和外极面图象方法，从而使之能适用于具有抖动的
图象序列分析，并避免了一般运动视觉方法的不适定问题和特征对应问题、基于
空域约束的迭代方法的局部最小化等问题。
3．提出了全方位图象特征和神经网络相结合的道路建模方法，较好地解决了
机器人依赖局部视野信息迷路的问题、视觉算法依赖特定环境特征推广性差等问
题。本文提出了主分量分析和 Fourier变换相结合的全方位图象数据压缩和旋转
不变特征提取方法，设计了在识别道路类型基础上进行道路方向估计的组合神经
元网络，从而提供了解决机器人在不同类型道路上自动切换和自适应道路跟踪的
可行方法。
4．提出了基于重投影变换的障碍物检测方法，跳出了传统立体视觉特征抽
取、匹配和三维恢复的模式。设计了无特征提取和对应的立体视觉新算法，给出
了克服摄象机俯仰影响的动态重投影变换算法，增强了系统在颠簸的道路上运行
的适应性。利用重投影变换后双目图象对上路面特征零视差的特性，通过障碍物
有无判断和障碍物三维测量的“分步渐进”过程，可高效、可靠地对路面障碍物
实时检测。
5．系统实现：门）实现了全景建模中的图象稳定、遮挡恢复和深度分层，从
而为全局定位的自然路标提取和真实环境再现的图象合成打下了基础；C）设计
和实现了适合于室外环境的全方位成象系统和单摄象机双目立体成象系统：m
实验验证了全方位道路图象的神经网络学习方法的有效性；w实现的实时障碍
物检测系统己经过大量的室外道路环境下障碍物检测的试验，结果说明系统具有
很强的实用性。
Visual navigation of a mobile robot in the natural environment is a difficult and
comprehensive subject which is related to almost every aspects of computer vision
researches. The fundamental tasks of visual navigation are composed of global
localization, road following and obstacle detection. Environment modeling is the
foundation of visual navigation. This dissertation is devoted to the deep and
systematic study on the visual environment modeling. The main contributions include:
1. A task-oriented, multi-scale and full-view visual modeling strategy is proposed
for the natural environment which combines the panoramic vision for scene modeling,
omni-directional vision for road understanding and binocular vision for obstacle
detection together. This approach overcomes the drawbacks of traditional visual
navigation methods that mainly depended on local and/or single view visual
information. In this direction sensor design, data processing and model representation
are closely explored.
2. A two stage method is presented for the 3D panoramic scene modeling from
vibrated image sequences which consists of (I) image stabilization by motion
filtering and (2) depth estimation and depth boundary localization. The two stage
method not only combines Zheng and Tsuji抯 panoramic image method with Bakers
epipolar plane image analysis, resulting the so called panoramic epipolar plane image
method, but also generalizes them to handle image sequence vibrations due to the un-
controllable fluctuation of the camera. The two stage method by-passes the
correspondence problem and ill-posed problem encountered in the general motion
analysis, and avoids the local minimum problem of the spatial-constrain-based
iteration method.
3. .A new road following approach . the Road Omni-View Image Neural
Networks (ROVINN). is proposed which combines the omni-directional image
sensing technique with neural networks. The ROVINN makes the robot never get lost
and enables it to learn from the road images. The ROVINN approach brings Yagi抯
COPIS to the outdoor road scene and provides a new solution from the CMU抯
ALVINN. Compact and rotation-invariant image features are extracted by integrating
the principle component analysis (PCA) and the Fourier transform (DFT). The
modular neural networks can estimate road orientations more efficiently by first
classif~抜ng the roads, and thus enable the robot to adapt to various road types
automatically.
4. A novel method called image reprojection transformation is presented for road
obstacle detection based on binocular vision. Dynamic reprojection transformation
algorithms are developed so as to work in un-even road surface. The novelty of the
(dynamic) reprojection transformation method, which ensembles the gaze control of
the human vision, lies in the fact that it brings the road surface to zero disparity so that
the feature extraction and matching procedures of the traditional stereo vision are
avoided in the obstacle detection task. The progressive processing strategy of
reproj ection transformation, yes/no verification and obstacle measurement make the
obstacle detection efficient, fast and robust.
5. System implementations.(l) In the 3D panoramic scene modeling , the
algorithms of motion filtering and image stabilization, kinetic occlusion detection and
depth layering, have been developed so as to found a ground base for landmark
selection of global localization and image synthesis of virtualized rea
难度的综合性课题。视觉导航的基本任务是全局定位、道路跟踪和障碍物检测，
其中每一部分都要以环境建模为基础。本文系统深入地研究了环境建模问题，包
括：
1．提出了面向任务的多尺度和全覆盖视野的环境综合建模方法。将路标定位
的全景图象建模，道路识别的全方位图象建模和目标（障碍物）检测的双目注视图
象建模相结合，改进了依赖局部、单一信息的视觉导航方法。研究了面向任务的
环境建模方法中的传感器设计、视觉算法和模型表示问题。
2．提出了由非精确摄象机运动下的图象序列建立3D环境全景模型的两步
法，即基于运动滤波的图象稳定和基于时- 空－频域遮挡模型的全景外极面图象分
析。本文推广了全景图象方法和外极面图象方法，从而使之能适用于具有抖动的
图象序列分析，并避免了一般运动视觉方法的不适定问题和特征对应问题、基于
空域约束的迭代方法的局部最小化等问题。
3．提出了全方位图象特征和神经网络相结合的道路建模方法，较好地解决了
机器人依赖局部视野信息迷路的问题、视觉算法依赖特定环境特征推广性差等问
题。本文提出了主分量分析和 Fourier变换相结合的全方位图象数据压缩和旋转
不变特征提取方法，设计了在识别道路类型基础上进行道路方向估计的组合神经
元网络，从而提供了解决机器人在不同类型道路上自动切换和自适应道路跟踪的
可行方法。
4．提出了基于重投影变换的障碍物检测方法，跳出了传统立体视觉特征抽
取、匹配和三维恢复的模式。设计了无特征提取和对应的立体视觉新算法，给出
了克服摄象机俯仰影响的动态重投影变换算法，增强了系统在颠簸的道路上运行
的适应性。利用重投影变换后双目图象对上路面特征零视差的特性，通过障碍物
有无判断和障碍物三维测量的“分步渐进”过程，可高效、可靠地对路面障碍物
实时检测。
5．系统实现：门）实现了全景建模中的图象稳定、遮挡恢复和深度分层，从
而为全局定位的自然路标提取和真实环境再现的图象合成打下了基础；C）设计
和实现了适合于室外环境的全方位成象系统和单摄象机双目立体成象系统：m
实验验证了全方位道路图象的神经网络学习方法的有效性；w实现的实时障碍
物检测系统己经过大量的室外道路环境下障碍物检测的试验，结果说明系统具有
很强的实用性。
Visual navigation of a mobile robot in the natural environment is a difficult and
comprehensive subject which is related to almost every aspects of computer vision
researches. The fundamental tasks of visual navigation are composed of global
localization, road following and obstacle detection. Environment modeling is the
foundation of visual navigation. This dissertation is devoted to the deep and
systematic study on the visual environment modeling. The main contributions include:
1. A task-oriented, multi-scale and full-view visual modeling strategy is proposed
for the natural environment which combines the panoramic vision for scene modeling,
omni-directional vision for road understanding and binocular vision for obstacle
detection together. This approach overcomes the drawbacks of traditional visual
navigation methods that mainly depended on local and/or single view visual
information. In this direction sensor design, data processing and model representation
are closely explored.
2. A two stage method is presented for the 3D panoramic scene modeling from
vibrated image sequences which consists of (I) image stabilization by motion
filtering and (2) depth estimation and depth boundary localization. The two stage
method not only combines Zheng and Tsuji抯 panoramic image method with Bakers
epipolar plane image analysis, resulting the so called panoramic epipolar plane image
method, but also generalizes them to handle image sequence vibrations due to the un-
controllable fluctuation of the camera. The two stage method by-passes the
correspondence problem and ill-posed problem encountered in the general motion
analysis, and avoids the local minimum problem of the spatial-constrain-based
iteration method.
3. .A new road following approach . the Road Omni-View Image Neural
Networks (ROVINN). is proposed which combines the omni-directional image
sensing technique with neural networks. The ROVINN makes the robot never get lost
and enables it to learn from the road images. The ROVINN approach brings Yagi抯
COPIS to the outdoor road scene and provides a new solution from the CMU抯
ALVINN. Compact and rotation-invariant image features are extracted by integrating
the principle component analysis (PCA) and the Fourier transform (DFT). The
modular neural networks can estimate road orientations more efficiently by first
classif~抜ng the roads, and thus enable the robot to adapt to various road types
automatically.
4. A novel method called image reprojection transformation is presented for road
obstacle detection based on binocular vision. Dynamic reprojection transformation
algorithms are developed so as to work in un-even road surface. The novelty of the
(dynamic) reprojection transformation method, which ensembles the gaze control of
the human vision, lies in the fact that it brings the road surface to zero disparity so that
the feature extraction and matching procedures of the traditional stereo vision are
avoided in the obstacle detection task. The progressive processing strategy of
reproj ection transformation, yes/no verification and obstacle measurement make the
obstacle detection efficient, fast and robust.
5. System implementations.(l) In the 3D panoramic scene modeling , the
algorithms of motion filtering and image stabilization, kinetic occlusion detection and
depth layering, have been developed so as to found a ground base for landmark
selection of global localization and image synthesis of virtualized rea
引文
1 Abbott A L, Zhang B, Active Fixation using attentional shifts, affine resampling and multi resolution search, Proc ICCV 1995:1002-1008
2 Abbott A L, Ahuja N, Surface reconstruction by dynamic integration of focus, camera vergence and stereo , Proc ICCV 1988:532-543
3 Adelson, E.H., Bergen, J.R., Spatiotemporal energy model for the perception of motion, J. Opt. Soc. Am.,A2,284-299,1985
4 Allmen M, Dyer C R, Computing spatiotemporal surface flow, Proc IEEE 3rd ICCV, 1990: p47-50
5 Allmen M, Dyer C R, Long range spatiotemporal motion understanding using spatiotemporal flow curves, Proc IEEE CVPR 1991: p303-309
6 Aloimonos Y, Weiss I, Bandopadhay A, Active vision, Int J CV, vol 2, 1988:333-356
7 Aloimonos Y, Purposive and qualitative active vision, Proc. IEEE 10th ICPR1990:346-360
8 Aloimonos Y, Rosenfeld A, Computer vision. Science, Vol253． Sept 1991:1181-1324．
9 Ariyaeeinia A M, Single camera 3D TV system, SPIE vol 2351 Telemanipulator and Telepresence Technologies, 1994:255-232
10 Atsumi E ,et al, Internal representation of a neural network that detects local motion , Proc. IJCNN, 1993: 198-201
11 Azarbayejani A , Penland A , Recursive estimation of motion, structure, and focal length, IEEE Trans PAMI, Vol. 17, No. 6, Jun. 1995, pp562-575．
12 Bachelder I A, Waxman A M, Mobile robot visual mapping and localization: a view-based neural computed architecture that emulates hippocampal place learning, Neural Networks, Vil7Nos6/7 1994:1083-1099
13 Badal S, Ravela S, Draper B, Hanson A, A practical obstacle detection and avoidance system, IEEE Workshop on Applications of Computer Vision, 1994．
14 BajscyR, Active perception, Proc IEEE, vol 76,1988:996-1005
15 Baker H H, Bolles R C, Generalizing epipolar-plane image analysis on the spatio-temporoal surface CVPR 1988,pp1510
16 Baker H H, Bolles R C, Generalization epipolar-plane image analysis on the spatio-temporal surface, Int. J. CV. 3． 1989:33-49
17 Ballard D H, Behavioral constraints on animate vision, Image and Vision Computing, 7(1)
FEB1989:3-9
18 Ballard D H, Barrow C M, Principles of animate vision, CVGIP:IU:56(1) July 1992:3-21
19 Bani-Hashemi A, A Fourier approach to camera orientation, In: Proc ICRA, Nice, France, May 1992, 1532-1538
20 Bhanu B, Nevatia R, Riseman E M, Dynamic scene and motion analysis using passive sensor, PART I: A qualitative approach Feb 1992 45-52, PART Ⅱ: Displacement-Field and feature based approaches53-64
21 Black M J, Jepson A D, Estimating optical flow in segmented images using variable-order parametric models with local deformations, IEEE Trans PAMI, vol 18, no 10, Oct 1996: 972-986．
22 Broida T S, Chellappa R, Estimation object motion parameters from noisy images, IEEE Trans PAMI, 8, 1986,:90-99
23 Brooks R A, Achieving AI through Building Robots, AI Memo 899, Al Lab, MIT, Cambridge MA, 1986
24 Burt P, Wixson L, Salgian G, Electronically directed "focal" stereo, ICCV 95:94-101
25 Cipolla R, Yamamoto R, Stereoscopic tracking of bodied in motion, Image and Vision Computing, vol 8, no 1, Feb 1990:85-90
26 Cho Y C, Cho H S, A Stereo Vision-based Obstacle Detection method for mobile robot navigation, Robotica 12,1994:203-216
27 Chou G T, A model of figure-ground segregation from kinetic occlusion, Pro ICCV, 1995:1050-1057．
28 Collins R, Janes C , Cheng Y Q, et al, The Umass ascender system for 3D site model construction, chapter to appear in forthcoming ARPA IU RADIUS book, Osacar Firshein (Ed.), 1996
29 Coombo, D., Brown, C., Real-time binocular smooth pursuit, Int J Computer Vision, vol 11, no 2． 1994:PP147-164．
30 Cortelazzo G and Balanza M , Frequency domain analysis of translations with piecewise cubic trajectories, IEEE Trans PAMI, vol 15． No 4, April 1993: p411-416
31 Costeira J, Kanade T, A multi-body factorization method for motion analysis, Proc ICCV, 1995: 1071-1076
32 Dalmia A K, Trivedi M, High speed extraction of 3D structure of selectable quality using a translating camera, CV1U vol. 64, No 1 .July 1996:97-110
33 Dickmanns F D, Graefe V, Dynamic Monocular Machine Vision, Int J MVA, Vol. 1． 1988:223-240
34 Dhond U R, Aggarwal J K, Structure from stereo-A review, IEEE Trans SMC, 19 1989 1489-1510
35 Elkins R T, Hall E L, Three dimensional line following using omnidirectional vision, Proc. SPIE vol. 2354 , 1994: pp 130-144
36 Fermiiller C, Aloimonos Y, Vision and action, Image and Vision Computing, Vol. 13, No 10, Dec 1995:725-744
37 Freeman W T, Adelson E H, The design and use of steerable filters, IEEE Trans PAMI vol 13, no 9, Sep 1991:891-906
38 Goshtasby A, Gruver W A, Design of a single-lens stereo camera system, Pattern Recognotion, vol 26, no 6, 1993:923-937．
39 Halfhill T R, See you around, BYTE,May 1995:85-90．
40 Heeger D J , Optical flow from spatio-temporal filters, In Proc. ICCV, 1987a: pl81-190
41 Heeger D J, Model for the extraction of image flow, J. Opt. Soc. Am. A, vol 4, no 8, August 1987b:1455-1471
42 Hong J, Tan X, Pinette B, et al, Image-based homing, Proc Int Conf Robotics and Automation, April 1991, pp 620-625．
43 Horn B K P, Robot Vision, The MIT Press, McGrawhill Book Company, 1986
44 Hsu S, Anandan P, Peleg S, Accurate computation of optical flow by using layered representation, Proc ICPR 1994: 743-746．
45 Irani M, Rousso B, Peleg S, Computing occluding and transparent motion, Int J CV, 12(1) ,1994:5-16
46 Irani M, Ananda P, Hsu S, Mosaic based representation of video sequence and their applications, IEEE Proc ICCV'95, pp605-611．
47 Ishiguro H, Yamamoto M, Tsuji S, Omni-directional stereo for making global map, Proc. ICCV-90．
48 Jahne B, Digital Image Processing , Concept, Algorithms and Scientific Applications, Springer-Verlag, 1991
49 Jochem T, Using virtual active tools to improve autonomous driving tasks, Technical Report, Nov, 1993．
50 Jochem T, Pomerleau D, Thorpe C, MANIC: A next generation neurally based autonomous road follower, Proc IU Workshop, Washington ,USA, April, 1993
51 Jochem T, Pomerleau D, Thorpe C, Vision guided lane transition, IEEE Sym on Intelligent Vehicles, Detroit, USA, Sep 25-26,1995
52 Kak, A. C.(ed.), Computer Vision and Image Understanding, special issue on . Functionality in Object Recognition, vol. 62, no,2, Sep 1995．
53 Kanatani K 1, Constraint on length and angle,CVGIP,41 1988:28-42
54 Kayaalp A , Eckman J L, Near real-time range detection using a pipeline architecture, IEEE Trans SMC, vol. 20, no 6, Nov/Dec 1990．
55 Kirby M, Sirovich L, Application of the Karhunen-Loeve procedure for the characterization of human faces, IEEE Trans PAMI vol 12, no 1, Jan, 1990．
56 Kirby M, Weisser F, Dangelmayr G, A model problem in the representation of digital image sequences, Pattern Recognition, vol 26, no 1, 1993: 63-73
57 Kumar R, Anandan P , Hanna K, Direct recovery of shape from multiple views: a parallax based approach. ICPR-94:685-688．
58 Kurada S, Rankin G W, Sridhar, A trinocular vision system for close-range postion sensing, Optics and Laser Technology, vol 27, no 2 1995:75-79．
59 Langley K, Fleet D, Atherton T, Multiple motions from instantaneous frequency, Proc Int Conf CVPR, 1992:846-849．
60 Li S , Tsuji S, Selecting distinctive scene features for landmarks, Proc ICRA, 1992:53-59
61 Li S , Nagata S, Tsuji S, A navigation system based upon panoramic representation, Proc IEEE/RSJIROS,1995 :p 142-7．
62 Lin. X.Y. , Zhu, Z.G., "Detecting Height From Constrained Motion", Proc. IEEE 3rd International Conference on Computer Vision, Osaka, Japan, 1990: 503-506
63 Lin X Y, Zhu Z G, Den W., "A Stereo Matching Algorithm Based on Shape Similarity for Indoor Environment Model Building," Proc. IEEE International Conference Robotics and Automation, 1996; 191 6-1922
64 Marr D, Vision: A Computational Investigation into the Human Representation and Processing of Visual Information, Freeman, San Francisco, 1982 中译本：姚国正，刘磊，汪云九译，计算视觉理论，科学出版社，1988。
65 Mann S, Picard R W, Video orbit of the projective group: a new perspective on image mosaicing. Technical Report No. 338, M.I.T. Media Lab Perceptual Computing Section. 1995．
66 Matthies L, Szeliski R Kanade T, Kalman filter-based algorithms for estimating depth from image sequence. Proc DARPA 1U Workshop. 1988:199-213
67 Murray D W, Recovering range using virtual multicamera stereo, CVIU, Vol. 61,2． March 1995:285-291
68 Murase H, Nyar S K, Visual learning and recognition of 3D objects from appearance, Int. J.
Computer Vision, 14,1995 :5-24
69 Nelson R C, Aloimonos Y, Obstacle avoidance using flow field divergence, IEEE PAMI-11(10) ,1989:1102-1106
70 Nelson R C, Polana R, Qualitative recognition of motion using temporal texture, CVGIP: Image Understanding, 1992,56(1) :78~89
71 Neuralware Inc., Neural Computing / Reference Guide/ Using Nworks of Professional Ⅱ/Plus, 1991．
72 Niyogi S A, Detecting kinetic occlusion, Proc ICCV, 1995:1044-1049
73 Nishimoto Y , Shirai Y, A feature-based stereo model using small disparities, Proc CVPR 1987:912-196
74 Okutomi M, Kanade T, A multiple-baseline stereo, Proc CVPR 1991:63-69
75 Poggio T, Girosi F, Networks for approximation and learning proceedings of the IEEE, vol 78, no 9, Sep. 1990．
76 Pomerleau D, Neural network based autonomous navigation, In Vision and Navigation : the CMU Navlab, Kluwer Academic Publishers, Boston, 1990．
77 Pomerleau D, PALPH: Rapidly Adapting Lateral Position Handler, IEEE Sym. on Intelligent Vehicles, Detroit, USA, Sep 25-26,1995
78 Polis M F, Gifford S J, Mckeown D M Jr., Automating the construction of large-scale virtual worlds, Computer, July, 1995:pp57-65．
79 Powell I, Panoramic lens, Applied Optics, vol 33, no 31, Nov 1994:7356-7361
80 Rivlin E, Rosenfeld A, Navigational functionalities, Computer Vision and Image Understanding, vol 62, no 2, Sep 1995: 232-244．
81 Rojer A, Schwartz E., Design consideration for a space-variant visual sensor with complex-logarithmic geometry, Proc. ICPR, June, 1990．
82 Sawhney H S, Simplifying motion and structure analysis using planar parallax and image warping, ICPR-94:403-408
83 Sawhney H S, Ayer S, Compact representation of videos through dominant and multiple motion estimation", IEEE Trans. PAMI, Vol. 18, No. 8, Aug 1996, pp814-830．
84 Sclaroff S, Pentland P, Modal matching for correspondence and recognition, IEEE Trans PAMI, vol 17, no 6, 1995 : 545-561
85 Simpson R L, Computer Vision : an overview, special issue on DARPA Vision and Navigation, August 1991:11-15
86 Shashua A, Navab N, Relative affine structure: canonical model for 3D from 2D geometry and applications, IEEE Trans PAMI vol 18, no9, Sep 1996: 873-883
87 Stein F , Medioni G, Map-based localization using the panoramic horizon, Proc. ICRA, 1992:2631-2637
88 Storjohann K , et al, Visual obstacle detection for automatically guided vehicles, Proc IEEE ICRA, pp761-766,1990．
89 Szeliski R, Image mosaicing for tele-reality application, Proc. 2nd IEEE Workshop on Application of Computer Vision, 1994, pp44-53．
90 Szeliski R, Shum H, Motion estimation with quadtree splines, IEEE Trans PAMI, vol 18, no 12, Dec:1996:pp1199-1210．
91 Taalebinezhaad M A, Direct recovery of motion and shape in the general case by fixation , IEEE PAMI-14(8) , Aug 1992:pp847-853
92 Teoh W, Zhang X D, An inexpensive stereoscopic vision system for robots, Proc Int Conf Robotics,1984:186-189．
93 Terzopoulos D, Regularization of inverse problems involving discontinuities ,IEEE Trans PAMI, 8,1986:413-424．
94 Thorpe C, et al. Vision and Navigation for the CMU Navlab IEEE Trans. PAMI, 1988, 10(3) : 362-376
95 Thorpe, C.(ed.) , Vision and Navigation : the CMU Navlab, Kluwer Academic Publishers, Boston, 1990．
96 Thorpe C, Hebert M, Kanade T, Shafer S, Toward autonomous Driving: The CMU Navlab, Vol. 6,No4 Aug 1991 31-52
97 Thrion J.-P., Realistic 3D simulation of shapes and shadows for image processing, CVGIP: Graphical Model and Image Processing, vol 54, No 1 , 1992 :p82-90．
98 Tomasi C, Kanade T, Shape and motion without depth, ICCV, 1990: 91-95
99 Tomasi C, Kanade T, Shape and motion from image streams under orthography: A factorization method, Int J CV, 9, 1992: 137-154
100 Ullman S, The Interpretation of structure from motion , Proc R, Soc London, Series B 203 1979,405-426
101 Wang J. Adelson E H. Representation moving images with layers. IEEE Trans, on Image Processing, Vol. 3, No. 5 Sep. 1994, pp625-638．
102 Weens C C, Brown C, Webb J A, Parallel Processing in the DARPA strategic Computing vision program, Oct. 1991: 23-38
103 Weisbin C R, Montemerlo M D, NASA's telerobotics research program, ICRA-.1992 : 2653-2666
104 Weiss Y, Adelson E H , Perceptually organized EM: a framework for motion segmentation that combines information about form and motion, MIT Media Lab Perceptual Computing Section Technical Report No. 315, 1995
105 Wong Q T, Weber J, Roller D, Malik J, A Integrated stereo-based approach to automatic vehicle guidance, ICCV:95:52-57
106 Yagi Y, Kawato S, Tsuji S, Collision avoidance using omnidirectional image sensor (COPIS), Proc IEEE ICRA, 1991
107 Yagi Y, Kawato S, Tsuji S, Real-time omnidirectional image sensor (COPIS) for vision-guided navigation, IEEE Trans Robotics and Automation, vol. 1, no 1, Feb 1994 :pp 11-27
108 Yagi Y, Okumara H, Yachida M, Multiple visual sensing system for mobile robot, Proc ICRA, 1994: 1679-84．
109 Yagi Y, Nishizawa Y, Yachida M, Map-based navigation for a mobile robot with omnidirectional image sensor COPIS, IEEE Trans Robotics and Automation, vol 11, no 5, 1995:634-648
110 Yagi Y, Yachida M, Evaluating effectivity of map generation by tracking vertical edges in omnidirectional image sequence, Proc ICRA,1995:2234-9．
111 Zhang Z, Weiss R, Hanson A R, Obstacle detection based on qualitative and quantitative 3D reconstruction, IEEE Trans PAMI, vol 19, no 1, Jan 1997: 15-26．
112 Zheng J Y, Barth M, Tsuji S, Panoramic representation of scenes for route understanding. In: Proc. IEEE 10th ICPR, 1990: pp 161-167
113 Zheng J.Y., Barth M, Tsuji S., "Autonomous landmark selection for route recognition by a mobile robot," Proc. ICRA, 1991
114 Zheng J Y, Tsuji S, Panoramic representation for route recognition by a mobile robot. Int J. CV, vol 9, no 1 1992: pp 55-76
115 (ZhengY90) Zheng Y, et a SWITCHER: A stereo algorithm for ground plane obstacle detection. Image and Vision Computing, 1990, 8 (1) : 57-62
116 Zhou Y T, Chellappa R A, A network for motion perception, Proc. IJCNN , 1990: pp Ⅱ841-851
117 Zhu Z G. Lin X Y, Realtime algorithms for obstacle avoidance by using reprojection transformation. In: Proc IAPR Workshop on Machine Vision Application, 1990． 393-396
118 Zhu Z G, Xu G Y , Chen S Y, Lin X Y, Dynamic obstacle detection through cooperation of purposive visual modules of color, stereo and motion, Proc IEEE ICRA, USA, May 1994．
119 Zhu Z G, Xu G Y, Neural networks for omni-view road image understanding, J. of Comput. Sci. and Technol. (计算机学报英文版), vol 11, no 4, July 1996a.
120 Zhu, Z.G., Yang, B., Xu. G.Y., VISATRAM: A visual system for automatic traffic monitoring based on 2D spatio-temporal images, Proc. IEEE Workshop on Applications of Computer Vision, Sarasota, FL, Dec. 2-4, 1996b: 162-167
121 包约翰著，马颂德，张恭清，高雨清译，神经元网络和自适应模式识别，科学出版社，1992
122 石定机，朱志刚，陈永康，立体象对实时拍摄装置，中国实用新型专利，ZL95211525．5，1995-12-9
123 邬永革，杨靖宇，基于证据理论处理移动机器人视觉信息的一种方法，计算机研究与发展，第31卷第12期，1994．12：pp：32-36
124 邬永革，刘雷健，杨靖宇，等，利用信息融合方法的三维特征提取，计算机研究与发展，第33卷第7期，1996．7：pp：528-532
125 杨雨东，徐光佑，朱志刚，2.5维帧间运动估计方法，清华大学学报(已录用，1997年刊出)．
126 曾建超 徐光佑 “虚拟现实技术及其发展战略”，电子学报，Vol．23 No．10 Oct．1995，pp57-61．
127 朱志刚，图象重投影变换原理及面向任务的运动分析算法，硕士论文，清华大学计算机系，1991．3
128 朱志刚，林学言，“图象重投影变换原理及其视觉应用”，《模式识别与人工智能》，第5卷，第2期，1992．6：pp155-162．
129 朱志刚，林学言，石定机，徐光佑，基于感知组织的图象2D推理和3D解释，电子学报，vol．23，no 11，1995．5．
2 Abbott A L, Ahuja N, Surface reconstruction by dynamic integration of focus, camera vergence and stereo , Proc ICCV 1988:532-543
3 Adelson, E.H., Bergen, J.R., Spatiotemporal energy model for the perception of motion, J. Opt. Soc. Am.,A2,284-299,1985
4 Allmen M, Dyer C R, Computing spatiotemporal surface flow, Proc IEEE 3rd ICCV, 1990: p47-50
5 Allmen M, Dyer C R, Long range spatiotemporal motion understanding using spatiotemporal flow curves, Proc IEEE CVPR 1991: p303-309
6 Aloimonos Y, Weiss I, Bandopadhay A, Active vision, Int J CV, vol 2, 1988:333-356
7 Aloimonos Y, Purposive and qualitative active vision, Proc. IEEE 10th ICPR1990:346-360
8 Aloimonos Y, Rosenfeld A, Computer vision. Science, Vol253． Sept 1991:1181-1324．
9 Ariyaeeinia A M, Single camera 3D TV system, SPIE vol 2351 Telemanipulator and Telepresence Technologies, 1994:255-232
10 Atsumi E ,et al, Internal representation of a neural network that detects local motion , Proc. IJCNN, 1993: 198-201
11 Azarbayejani A , Penland A , Recursive estimation of motion, structure, and focal length, IEEE Trans PAMI, Vol. 17, No. 6, Jun. 1995, pp562-575．
12 Bachelder I A, Waxman A M, Mobile robot visual mapping and localization: a view-based neural computed architecture that emulates hippocampal place learning, Neural Networks, Vil7Nos6/7 1994:1083-1099
13 Badal S, Ravela S, Draper B, Hanson A, A practical obstacle detection and avoidance system, IEEE Workshop on Applications of Computer Vision, 1994．
14 BajscyR, Active perception, Proc IEEE, vol 76,1988:996-1005
15 Baker H H, Bolles R C, Generalizing epipolar-plane image analysis on the spatio-temporoal surface CVPR 1988,pp1510
16 Baker H H, Bolles R C, Generalization epipolar-plane image analysis on the spatio-temporal surface, Int. J. CV. 3． 1989:33-49
17 Ballard D H, Behavioral constraints on animate vision, Image and Vision Computing, 7(1)
FEB1989:3-9
18 Ballard D H, Barrow C M, Principles of animate vision, CVGIP:IU:56(1) July 1992:3-21
19 Bani-Hashemi A, A Fourier approach to camera orientation, In: Proc ICRA, Nice, France, May 1992, 1532-1538
20 Bhanu B, Nevatia R, Riseman E M, Dynamic scene and motion analysis using passive sensor, PART I: A qualitative approach Feb 1992 45-52, PART Ⅱ: Displacement-Field and feature based approaches53-64
21 Black M J, Jepson A D, Estimating optical flow in segmented images using variable-order parametric models with local deformations, IEEE Trans PAMI, vol 18, no 10, Oct 1996: 972-986．
22 Broida T S, Chellappa R, Estimation object motion parameters from noisy images, IEEE Trans PAMI, 8, 1986,:90-99
23 Brooks R A, Achieving AI through Building Robots, AI Memo 899, Al Lab, MIT, Cambridge MA, 1986
24 Burt P, Wixson L, Salgian G, Electronically directed "focal" stereo, ICCV 95:94-101
25 Cipolla R, Yamamoto R, Stereoscopic tracking of bodied in motion, Image and Vision Computing, vol 8, no 1, Feb 1990:85-90
26 Cho Y C, Cho H S, A Stereo Vision-based Obstacle Detection method for mobile robot navigation, Robotica 12,1994:203-216
27 Chou G T, A model of figure-ground segregation from kinetic occlusion, Pro ICCV, 1995:1050-1057．
28 Collins R, Janes C , Cheng Y Q, et al, The Umass ascender system for 3D site model construction, chapter to appear in forthcoming ARPA IU RADIUS book, Osacar Firshein (Ed.), 1996
29 Coombo, D., Brown, C., Real-time binocular smooth pursuit, Int J Computer Vision, vol 11, no 2． 1994:PP147-164．
30 Cortelazzo G and Balanza M , Frequency domain analysis of translations with piecewise cubic trajectories, IEEE Trans PAMI, vol 15． No 4, April 1993: p411-416
31 Costeira J, Kanade T, A multi-body factorization method for motion analysis, Proc ICCV, 1995: 1071-1076
32 Dalmia A K, Trivedi M, High speed extraction of 3D structure of selectable quality using a translating camera, CV1U vol. 64, No 1 .July 1996:97-110
33 Dickmanns F D, Graefe V, Dynamic Monocular Machine Vision, Int J MVA, Vol. 1． 1988:223-240
34 Dhond U R, Aggarwal J K, Structure from stereo-A review, IEEE Trans SMC, 19 1989 1489-1510
35 Elkins R T, Hall E L, Three dimensional line following using omnidirectional vision, Proc. SPIE vol. 2354 , 1994: pp 130-144
36 Fermiiller C, Aloimonos Y, Vision and action, Image and Vision Computing, Vol. 13, No 10, Dec 1995:725-744
37 Freeman W T, Adelson E H, The design and use of steerable filters, IEEE Trans PAMI vol 13, no 9, Sep 1991:891-906
38 Goshtasby A, Gruver W A, Design of a single-lens stereo camera system, Pattern Recognotion, vol 26, no 6, 1993:923-937．
39 Halfhill T R, See you around, BYTE,May 1995:85-90．
40 Heeger D J , Optical flow from spatio-temporal filters, In Proc. ICCV, 1987a: pl81-190
41 Heeger D J, Model for the extraction of image flow, J. Opt. Soc. Am. A, vol 4, no 8, August 1987b:1455-1471
42 Hong J, Tan X, Pinette B, et al, Image-based homing, Proc Int Conf Robotics and Automation, April 1991, pp 620-625．
43 Horn B K P, Robot Vision, The MIT Press, McGrawhill Book Company, 1986
44 Hsu S, Anandan P, Peleg S, Accurate computation of optical flow by using layered representation, Proc ICPR 1994: 743-746．
45 Irani M, Rousso B, Peleg S, Computing occluding and transparent motion, Int J CV, 12(1) ,1994:5-16
46 Irani M, Ananda P, Hsu S, Mosaic based representation of video sequence and their applications, IEEE Proc ICCV'95, pp605-611．
47 Ishiguro H, Yamamoto M, Tsuji S, Omni-directional stereo for making global map, Proc. ICCV-90．
48 Jahne B, Digital Image Processing , Concept, Algorithms and Scientific Applications, Springer-Verlag, 1991
49 Jochem T, Using virtual active tools to improve autonomous driving tasks, Technical Report, Nov, 1993．
50 Jochem T, Pomerleau D, Thorpe C, MANIC: A next generation neurally based autonomous road follower, Proc IU Workshop, Washington ,USA, April, 1993
51 Jochem T, Pomerleau D, Thorpe C, Vision guided lane transition, IEEE Sym on Intelligent Vehicles, Detroit, USA, Sep 25-26,1995
52 Kak, A. C.(ed.), Computer Vision and Image Understanding, special issue on . Functionality in Object Recognition, vol. 62, no,2, Sep 1995．
53 Kanatani K 1, Constraint on length and angle,CVGIP,41 1988:28-42
54 Kayaalp A , Eckman J L, Near real-time range detection using a pipeline architecture, IEEE Trans SMC, vol. 20, no 6, Nov/Dec 1990．
55 Kirby M, Sirovich L, Application of the Karhunen-Loeve procedure for the characterization of human faces, IEEE Trans PAMI vol 12, no 1, Jan, 1990．
56 Kirby M, Weisser F, Dangelmayr G, A model problem in the representation of digital image sequences, Pattern Recognition, vol 26, no 1, 1993: 63-73
57 Kumar R, Anandan P , Hanna K, Direct recovery of shape from multiple views: a parallax based approach. ICPR-94:685-688．
58 Kurada S, Rankin G W, Sridhar, A trinocular vision system for close-range postion sensing, Optics and Laser Technology, vol 27, no 2 1995:75-79．
59 Langley K, Fleet D, Atherton T, Multiple motions from instantaneous frequency, Proc Int Conf CVPR, 1992:846-849．
60 Li S , Tsuji S, Selecting distinctive scene features for landmarks, Proc ICRA, 1992:53-59
61 Li S , Nagata S, Tsuji S, A navigation system based upon panoramic representation, Proc IEEE/RSJIROS,1995 :p 142-7．
62 Lin. X.Y. , Zhu, Z.G., "Detecting Height From Constrained Motion", Proc. IEEE 3rd International Conference on Computer Vision, Osaka, Japan, 1990: 503-506
63 Lin X Y, Zhu Z G, Den W., "A Stereo Matching Algorithm Based on Shape Similarity for Indoor Environment Model Building," Proc. IEEE International Conference Robotics and Automation, 1996; 191 6-1922
64 Marr D, Vision: A Computational Investigation into the Human Representation and Processing of Visual Information, Freeman, San Francisco, 1982 中译本：姚国正，刘磊，汪云九译，计算视觉理论，科学出版社，1988。
65 Mann S, Picard R W, Video orbit of the projective group: a new perspective on image mosaicing. Technical Report No. 338, M.I.T. Media Lab Perceptual Computing Section. 1995．
66 Matthies L, Szeliski R Kanade T, Kalman filter-based algorithms for estimating depth from image sequence. Proc DARPA 1U Workshop. 1988:199-213
67 Murray D W, Recovering range using virtual multicamera stereo, CVIU, Vol. 61,2． March 1995:285-291
68 Murase H, Nyar S K, Visual learning and recognition of 3D objects from appearance, Int. J.
Computer Vision, 14,1995 :5-24
69 Nelson R C, Aloimonos Y, Obstacle avoidance using flow field divergence, IEEE PAMI-11(10) ,1989:1102-1106
70 Nelson R C, Polana R, Qualitative recognition of motion using temporal texture, CVGIP: Image Understanding, 1992,56(1) :78~89
71 Neuralware Inc., Neural Computing / Reference Guide/ Using Nworks of Professional Ⅱ/Plus, 1991．
72 Niyogi S A, Detecting kinetic occlusion, Proc ICCV, 1995:1044-1049
73 Nishimoto Y , Shirai Y, A feature-based stereo model using small disparities, Proc CVPR 1987:912-196
74 Okutomi M, Kanade T, A multiple-baseline stereo, Proc CVPR 1991:63-69
75 Poggio T, Girosi F, Networks for approximation and learning proceedings of the IEEE, vol 78, no 9, Sep. 1990．
76 Pomerleau D, Neural network based autonomous navigation, In Vision and Navigation : the CMU Navlab, Kluwer Academic Publishers, Boston, 1990．
77 Pomerleau D, PALPH: Rapidly Adapting Lateral Position Handler, IEEE Sym. on Intelligent Vehicles, Detroit, USA, Sep 25-26,1995
78 Polis M F, Gifford S J, Mckeown D M Jr., Automating the construction of large-scale virtual worlds, Computer, July, 1995:pp57-65．
79 Powell I, Panoramic lens, Applied Optics, vol 33, no 31, Nov 1994:7356-7361
80 Rivlin E, Rosenfeld A, Navigational functionalities, Computer Vision and Image Understanding, vol 62, no 2, Sep 1995: 232-244．
81 Rojer A, Schwartz E., Design consideration for a space-variant visual sensor with complex-logarithmic geometry, Proc. ICPR, June, 1990．
82 Sawhney H S, Simplifying motion and structure analysis using planar parallax and image warping, ICPR-94:403-408
83 Sawhney H S, Ayer S, Compact representation of videos through dominant and multiple motion estimation", IEEE Trans. PAMI, Vol. 18, No. 8, Aug 1996, pp814-830．
84 Sclaroff S, Pentland P, Modal matching for correspondence and recognition, IEEE Trans PAMI, vol 17, no 6, 1995 : 545-561
85 Simpson R L, Computer Vision : an overview, special issue on DARPA Vision and Navigation, August 1991:11-15
86 Shashua A, Navab N, Relative affine structure: canonical model for 3D from 2D geometry and applications, IEEE Trans PAMI vol 18, no9, Sep 1996: 873-883
87 Stein F , Medioni G, Map-based localization using the panoramic horizon, Proc. ICRA, 1992:2631-2637
88 Storjohann K , et al, Visual obstacle detection for automatically guided vehicles, Proc IEEE ICRA, pp761-766,1990．
89 Szeliski R, Image mosaicing for tele-reality application, Proc. 2nd IEEE Workshop on Application of Computer Vision, 1994, pp44-53．
90 Szeliski R, Shum H, Motion estimation with quadtree splines, IEEE Trans PAMI, vol 18, no 12, Dec:1996:pp1199-1210．
91 Taalebinezhaad M A, Direct recovery of motion and shape in the general case by fixation , IEEE PAMI-14(8) , Aug 1992:pp847-853
92 Teoh W, Zhang X D, An inexpensive stereoscopic vision system for robots, Proc Int Conf Robotics,1984:186-189．
93 Terzopoulos D, Regularization of inverse problems involving discontinuities ,IEEE Trans PAMI, 8,1986:413-424．
94 Thorpe C, et al. Vision and Navigation for the CMU Navlab IEEE Trans. PAMI, 1988, 10(3) : 362-376
95 Thorpe, C.(ed.) , Vision and Navigation : the CMU Navlab, Kluwer Academic Publishers, Boston, 1990．
96 Thorpe C, Hebert M, Kanade T, Shafer S, Toward autonomous Driving: The CMU Navlab, Vol. 6,No4 Aug 1991 31-52
97 Thrion J.-P., Realistic 3D simulation of shapes and shadows for image processing, CVGIP: Graphical Model and Image Processing, vol 54, No 1 , 1992 :p82-90．
98 Tomasi C, Kanade T, Shape and motion without depth, ICCV, 1990: 91-95
99 Tomasi C, Kanade T, Shape and motion from image streams under orthography: A factorization method, Int J CV, 9, 1992: 137-154
100 Ullman S, The Interpretation of structure from motion , Proc R, Soc London, Series B 203 1979,405-426
101 Wang J. Adelson E H. Representation moving images with layers. IEEE Trans, on Image Processing, Vol. 3, No. 5 Sep. 1994, pp625-638．
102 Weens C C, Brown C, Webb J A, Parallel Processing in the DARPA strategic Computing vision program, Oct. 1991: 23-38
103 Weisbin C R, Montemerlo M D, NASA's telerobotics research program, ICRA-.1992 : 2653-2666
104 Weiss Y, Adelson E H , Perceptually organized EM: a framework for motion segmentation that combines information about form and motion, MIT Media Lab Perceptual Computing Section Technical Report No. 315, 1995
105 Wong Q T, Weber J, Roller D, Malik J, A Integrated stereo-based approach to automatic vehicle guidance, ICCV:95:52-57
106 Yagi Y, Kawato S, Tsuji S, Collision avoidance using omnidirectional image sensor (COPIS), Proc IEEE ICRA, 1991
107 Yagi Y, Kawato S, Tsuji S, Real-time omnidirectional image sensor (COPIS) for vision-guided navigation, IEEE Trans Robotics and Automation, vol. 1, no 1, Feb 1994 :pp 11-27
108 Yagi Y, Okumara H, Yachida M, Multiple visual sensing system for mobile robot, Proc ICRA, 1994: 1679-84．
109 Yagi Y, Nishizawa Y, Yachida M, Map-based navigation for a mobile robot with omnidirectional image sensor COPIS, IEEE Trans Robotics and Automation, vol 11, no 5, 1995:634-648
110 Yagi Y, Yachida M, Evaluating effectivity of map generation by tracking vertical edges in omnidirectional image sequence, Proc ICRA,1995:2234-9．
111 Zhang Z, Weiss R, Hanson A R, Obstacle detection based on qualitative and quantitative 3D reconstruction, IEEE Trans PAMI, vol 19, no 1, Jan 1997: 15-26．
112 Zheng J Y, Barth M, Tsuji S, Panoramic representation of scenes for route understanding. In: Proc. IEEE 10th ICPR, 1990: pp 161-167
113 Zheng J.Y., Barth M, Tsuji S., "Autonomous landmark selection for route recognition by a mobile robot," Proc. ICRA, 1991
114 Zheng J Y, Tsuji S, Panoramic representation for route recognition by a mobile robot. Int J. CV, vol 9, no 1 1992: pp 55-76
115 (ZhengY90) Zheng Y, et a SWITCHER: A stereo algorithm for ground plane obstacle detection. Image and Vision Computing, 1990, 8 (1) : 57-62
116 Zhou Y T, Chellappa R A, A network for motion perception, Proc. IJCNN , 1990: pp Ⅱ841-851
117 Zhu Z G. Lin X Y, Realtime algorithms for obstacle avoidance by using reprojection transformation. In: Proc IAPR Workshop on Machine Vision Application, 1990． 393-396
118 Zhu Z G, Xu G Y , Chen S Y, Lin X Y, Dynamic obstacle detection through cooperation of purposive visual modules of color, stereo and motion, Proc IEEE ICRA, USA, May 1994．
119 Zhu Z G, Xu G Y, Neural networks for omni-view road image understanding, J. of Comput. Sci. and Technol. (计算机学报英文版), vol 11, no 4, July 1996a.
120 Zhu, Z.G., Yang, B., Xu. G.Y., VISATRAM: A visual system for automatic traffic monitoring based on 2D spatio-temporal images, Proc. IEEE Workshop on Applications of Computer Vision, Sarasota, FL, Dec. 2-4, 1996b: 162-167
121 包约翰著，马颂德，张恭清，高雨清译，神经元网络和自适应模式识别，科学出版社，1992
122 石定机，朱志刚，陈永康，立体象对实时拍摄装置，中国实用新型专利，ZL95211525．5，1995-12-9
123 邬永革，杨靖宇，基于证据理论处理移动机器人视觉信息的一种方法，计算机研究与发展，第31卷第12期，1994．12：pp：32-36
124 邬永革，刘雷健，杨靖宇，等，利用信息融合方法的三维特征提取，计算机研究与发展，第33卷第7期，1996．7：pp：528-532
125 杨雨东，徐光佑，朱志刚，2.5维帧间运动估计方法，清华大学学报(已录用，1997年刊出)．
126 曾建超 徐光佑 “虚拟现实技术及其发展战略”，电子学报，Vol．23 No．10 Oct．1995，pp57-61．
127 朱志刚，图象重投影变换原理及面向任务的运动分析算法，硕士论文，清华大学计算机系，1991．3
128 朱志刚，林学言，“图象重投影变换原理及其视觉应用”，《模式识别与人工智能》，第5卷，第2期，1992．6：pp155-162．
129 朱志刚，林学言，石定机，徐光佑，基于感知组织的图象2D推理和3D解释，电子学报，vol．23，no 11，1995．5．