Post on 13-Jan-2016
description
CAP 5415 Computer Vision
General
Instructor: Dr. Alper Yilmaz Email: yilmaz@cs.ucf.edu Office: CSB 250
Course Office hours
Monday 3.00pm-4.30pm Wednesday 3.00pm-4.30pm
Grading Midterms (2) 20% Final 30% Assignments (bi-weekly) %20 Projects (2) %30
Class notes “Fundamentals of computer vision”, Dr. Mubarak Shah
available on web page (http://www.cs.ucf.edu/courses/cap6411/book.pdf)
Text book “Introductory Techniques for 3D Computer Vision” Trucco
and Verri, Prentice Hall.
Topics We’ll Cover Imaging Geometry Camera Modeling and Calibration Filtering and Enhancing Images Region Segmentation Color and Texture Line and Curve Detection Shape Analysis Stereo Motion and Optical Flow Structure from X
We may change order
Computer Vision
Image Analysis Single frame Image Understanding
Video Analysis Multiple frames, temporal information Video Understanding
Generating an Image
3D Scene Surface reflectance Surface structure (shape)
Light source Camera
Perspective Projection
Image
pla
ne
3D world
Pin
hole
ca
mera
Z
f
(X,Y,Z)
y
Z
Yfy
Z
Xfx
Orthographic Projection
Image
pla
ne
3D world
Z
(X,Y,Z)y
Yy
Xx
Discrete Domain Image Set of integer values in two
dimensions 0-255
Gray level image (1 matrix 0: Black 255: White
Color image (3 matrices) Red, Green, Blue
Resolution Number of rows, number of columns
Video
Sequence of image Frames per second Gray level video Color video
Digitization
Analog camera Analog to Digital converter Frame grabber
Digital camera Already digitized MPEG or JPEG
Image Formats
TIFF JPEG PGM, PPM PNM BMP MPEG Quick Time…
Computer Vision
Shape From “X” Shading (single image) Texture (single image) Stereo (two images) Motion (multiple images)
Stereo
http://www.vision3D.com/stereo.html
Example
Stereo Pair
Stereo Fun
candy dinosaur shark
Shape from Shading
Lambertian Surface Model
lightsource
surfacenormal
),,).(,,(.),( zyxzyx lllnnnLNyxf
),,)).(1,,(1
1(.),(
22 zyx lllqpqp
LNyxf
Sphere Example
),,(1
),,(
222
zyxR
nnn
z
y
y
zq
z
x
x
zp
yxRz
zyx
Vase Image
(1, 0, 1) (-1, 1, 1) (-1,-1, 1)
Object Motion
VideoSequence of images
An Image from Hamburg Taxi Sequence
Video Mosaics
Vision Lab. Sequence
Beach Volley Sequence
Sprite
Tracking in Multiple Cameras
Find field of view (FOV) lines Detect objects Associate objects
Story Segmentation
Explosion/fire Detection