Computer Vision: (An Introduction)
Computer Vision: (An Introduction)
(an introduction)
Dr Faisal Mufti ECE 7259 Advanced Computer Vision Spring 2012
Press, R. Hartley and A. Zisserman An invitation to 3D vision, Y. Ma, S. Soatto, J. Kosecka, and. S.S. Sastry, Springer-Verlag. Lecture notes
Web resources, Journals/Conferences
IEEE Trans Pattern Analysis and Machine Intelligence CVPR, ICIP
Those who do not send their contact address on time will not be added to the contact list!!!
Midterm Project
Synopsis Mid review Final presentation
Project report
Final Exam-
25%
Midterm Project
Synopsis Mid review
Attitude = 100%
Final presentation
Project report
5% 5% 7.5% 7.5%
Final Exam-
25%
Pre-requisites
undergraduate level Understanding of vectors and spaces and geometry is desirable. Everyone should have access to MATLAB
Vision Coupled with assignments and projects for hands-on experience and reinforcement of the concepts
Follow-up courses
Machine Vision, Pattern recognition, Machine learning Multimedia communications and Visual Computing
Vision is the primary sense by which humans perceive the world. Understanding is much more than simply seeing, e.g. the 3D magic eye illusions that appear nonsensical at first. Computer vision aims to bestow this perceptive ability on computers, via intelligent interpretation of visual data.
10
3D Magic Eye
11
3D Magic Eye
12
13
14
Complexity of environment
Base level human vision is too
good and it is backed by a perceptual engine that cannot be matched in an engineered system.
15
Blind-Spot Experiment
Draw an image similar to that below on a piece of paper (the dot and cross are about 6 inches apart)
Close your right eye and focus on the cross with your left eye Hold the image about 20 inches away from your face and move it slowly towards you The dot should disappear!
16
17
digital images are ideal for this. 2D image processing techniques for analysing these signals. 3D projective geometry for building a 3D model of the world from our data. Statistical tools to make decisions with confidence bounds. Algorithms to put it all together.
18
19
20
21
22
Sharpening
An image can be sharpened
23
Masked Blur
Using masks you can specify which part of an image to operate on. Eg blurring the background (using a simple linear filter)
24
Gradient of an Image
This gives the 'embossed' effect in Photoshop
25
26
27
Edge Detection
Detecting edge features in images.
This example uses Canny's method
28
Feature Extraction
Edges
Corner
Invariant Features
29
30
31
Skin Color
32
Segmentation
33
Colour Segmentation
An example of skin detection using a skin chrominance model
34
35
Periodic noise can be easily identified and attenuated in the frequency domain.
36
Reconstruction of 3D structure
Projective geometry is used to reconstruct the 3D structure of a scene from 2 or more views.
camera models, epipolar geometry
Fundamental matrix
3D reconstruction of a scene texture morphing
37
38
39
Projection Models Intrinsic (lens) Paramters Extrinsic (pose) Parameters Camera Calibration
40
Rigid, Similarity, Affine, & Projective Mappings Homography Estimation Image Warping
41
42
Stereo Vision
Stereo Camera Setups Stereo Disparity / Parallax Epipolar Geometry Correspondence Matching Triangulation / Depth Recovery
43
Stereo Cameras
and consider stereo cameras.
44
45
46
47
48
49
50
51
52
53
54
Camera Motion
Motion Field vs Optic Flow Flow Estimation Egomotion Estimation Structure from Motion
55
56
Video Tracking
58
59
Face Detection
60
Face Recognition
61
3D Model Generation
62
63
64
Camera Servoing
65
66
Approach: control pan, tilt and zoom of multiple cameras to acquire centered, synchronized video views of a moving person.
67
68
69
EyeVision Examples
70
71
Vehicle Tracking
72
Robot Perception
73
74
75
76
Augmented Reality
77
Augmented Reality