Computer Vision, Lecture 1
http://www.ugrad.cs.jhu.edu/~cs461
Professor Hager
http://www.cs.jhu.edu/~hager

Outline for Today
Outline and Organization of the course
What is Computer Vision
Some Applications of Computer Vision

What is Computer Vision?
Trucco and Verri
computing properties of the 3D world from one or more digital image
Stockman and Shapiro
To make useful decisions about real physical objects and scenes based on sensed images
Ballard and Brown
The construction of explicit, meaningful description of physical objects from images

Some Terms
Image processing:  the study of the properties of operators that produce images from other images
we will touch on image filtering and related operators from image processing
Machine Vision: a somewhat outdated term which now tends to refer to industrial vision applications where (usually) a single camera is used to solve a structured inspection task
           (demonstration of NI Vision Builder example)
Pattern Recognition: typically refers to the recognition of structures in 2D images (usually without reference to any underlying 3D information).
Photogrammetry: the science of measurement though non-contact sensing, e.g. terrain maps from satellite images.  Usually is  more focussed on accuracy issues.

Our Data Structure

What Information is in Images?

What Information is in Images?

Problems of Computer Vision: Modeling

Problems of Computer Vision: Modeling

Computer Vision vs. Graphics
Computer Graphics
Produce “plausible” images
You choose the models, conditions, imaging parameters, etc.
Computer Vision
Given real images with noise, sampling artifacts …
Estimate physically quantities
Ill-posed ---- what is the minimum world knowledge we need?

Problems of Computer Vision: Feature Extraction

Slide 12

Slide 13

Computer Vision vs. Image Processing
Image Processing
Mostly concerned with image-to-image transformations
Filtering
Enhancement
Compression
Computer Vision
Concerned with how images reflect the 3D world
Filtering for feature extraction
Enhancement for recognition/detection
Compression that preserves geometric information in images

Problems of Computer Vision: Segmentation and Grouping

Computer Vision vs. Human Vision

Illusions: What Do They Tell Us?

Illusions: What Do They Tell Us?

Illusions: What Do They Tell Us?

Illusions: What Do They Tell Us?

Illusions: What Do They Tell Us?

Illusions: What Do They Tell Us?

The Marr Paradigm
The goal of computer vision: to produce a “2 ½ D” sketch
Shape from stereo
Shape from shading
Shape from motion
.
.
.
.

Problems of Computer Vision: Stereo Vision

Random Dot StereoGram

THE ORGANIZATION OF AN IMAGE SEQUENCE

THE MOTION FIELD

MOVING CAMERAS ARE LIKE STEREO

THE EPIPOLAR CONSTRAINT

An Example
(Courtesy Carlo Tomasi)

Problems of Computer Vision: Recognition

Problems of Computer Vision: Recognition

Applications of Computer Vision: Biometrics
Face recognition
Iris scanning
Fingerprint recognition
Activity recognition

Applications of Computer Vision: Medical Imaging

Applications of Computer Vision: Medical Imaging

Applications of Computer Vision: HCI

Applications of Computer Vision: Image Databases
(Courtesy D. Forsyth & J. Ponce)

Applications of Computer Vision: Data Acquisition
(Jitendra Malik, Berkeley)

Applications of Computer Vision: Motion Control

Applications of Computer Vision: Motion Control
(CMU)

Applications of Computer Vision: Motion Control

Course Information
Use the course WEB site
http://www.ugrad.cs.jhu.edu/~cs461
For information on this course

Computer Vision, Lecture 2
http://www.ugrad.cs.jhu.edu/~cs461
Professor Hager
http://www.cs.jhu.edu/~hager

A Modern Digital Camera

How Cameras Produce Images
Basic process:
photons hit a detector
the detector becomes charged
the charge is read out as brightness
Sensor types:
CCD (charge-coupled device)
most common
high sensitivity
high power
cannot be individually addressed
blooming
CMOS
simple to fabricate (cheap)
lower sensitivity, lower power
can be individually addressed

What’s under the Hood

THE ORGANIZATION OF A 2D IMAGE

Digitization Effects
The “diameter” d of a pixel determines the highest frequency representable in an image
Real scenes may contain higher frequencies resulting in aliasing of the signal.
In practice, this effect is often dominated by other digitization artifacts.
One problem in particular is differing sampling rates between digitizer and camera readout of a row.

Storing Images
Non-lossy schemes
pbm/pgm/ppm/pnm
code for file type, size, number of bands, and maximum brightness
tif (lossless and lossy versions)
bmp
gif
Lossy schemes
jpg
uses Y Cb Cr color representation; subsamples the color
Uses DCT on result
Uses the fact the human system is less sensitive to color than spatial detail

GIF IMAGE FORMAT
GIF (Graphics Interchange Format)
Limited to 8 bits/pixel for both color and gray-scale.

TIFF IMAGE FORMAT
TIFF (Tagged Image File Format)
More general than GIF
Allows 24 bits/pixel
Supports 5 types of image compression including:
 RLE (Run length encoding)
LZW (Lempel-Ziv-Welch)
JPEG (Joint Photographic Experts Group)

A Modern Digital Camera (Firewire)

THE ORGANIZATION OF AN IMAGE SEQUENCE

BANDWIDTH REQUIREMENTS

Color

Slide 56

Region Tracking Video

Light Spectra

Sunlight

Slide 60

Slide 61

Color receptors

How Color Cameras Work
1 CCD cameras
A Bayer pattern is placed in front of the CCD
A Demosaicing process reads the pixels in a region and computes color and intensity
3 CCD camera use a beam splitter and 3 separate CCDs
higher color fidelity
needs lots of light
requires careful alignment of ccds

Slide 64

Standard Color System

Filtering Colors

METAMERISM
Two different Spectral Energy Distributions with the same RED, GREEN, BLUE response are termed metamers.

Slide 68

Slide 69

Homogeneous Region: Photometry

Homogeneous Color Region: Photometry

Slide 72

Slide 73

Slide 74

Slide 75

Slide 76

Slide 77