Introduction

• What IS computer vision?

the analysis of digital images by a computer

• Where do images come from?

1
 Applications
• Medical Imaging

CT image of a
patient’s abdomen
liver

kidney
kidney

2
Visible Man Slice Through Lung

3
3D Reconstruction of the Blood Vessel Tree

4
CBIR of Mouse Eye Images for Genetic Studies

5
Robotics
• 2D Gray-tone or Color Images

“Mars” rover

• 3D Range Images

What am I?

6
Image Databases:

Images from my Ground-Truth collection.

What categories of image databases exist today?

7
Abstract Regions for Object Recognition

Original Images Color Regions Texture Regions Line Clusters

8
Documents:

9
10
Insect Recognition for Ecology

11
Surveillance: Object and Event Recognition in Aerial Videos

Original Video Frame

12
Color Regions Structure Regions
Video Analysis

What are the objects? What are the events?

13
 Vision for Graphics:

Recent work of Steve Seitz (CSE) and Aaron Hertzman

on Computing the Geometry of Objects from Images.

14
Digital Image Terminology:

0 0 0 0 1 0 0
pixel (with value 94)
0 0 1 1 1 0 0
0 1 95 96 94 93 92 its 3x3 neighborhood
0 0 92 93 93 92 92
0 0 93 93 94 92 93 region of medium
0 1 92 93 93 93 93 intensity
0 0 94 95 95 96 95
resolution (7x7)

• binary image
• gray-scale (or gray-tone) image
• color image
• multi-spectral image
• range image
• labeled image 15
Goals of Image and Video Analysis

• Segment an image into useful regions

• Perform measurements on certain areas

• Determine what object(s) are in the scene liver

kidney spleen

• Calculate the precise location(s) of objects

• Visually inspect a manufactured object

• Construct a 3D model of the imaged object

• Find “interesting” events in a video

16
•The Three Stages of Computer Vision

• low-level

image image

• mid-level

image features

• high-level

features analysis

17
Low-Level

sharpening

blurring

18
Low-Level

Canny

original image edge image

Mid-Level

ORT

data
structure
circular arcs and line segments 19
edge image
Mid-level

K-means
clustering
(followed by
connected
component
analysis)

original color image regions of homogeneous color

data
structure
20
Low- to High-Level

low-level
edge image

mid-level

consistent
high-level line clusters

Building Recognition
21
Imaging and Image Representation

 Sensing Process
 Typical Sensing Devices
 Problems with Digital Images
 Image Formats
 Relationship of 3D Scenes to 2D Images
 Other Types of Sensors

22
Images: 2D projections of 3D
 The 3D world has color, texture, surfaces,
volumes, light sources, objects, motion, …
 A 2D image is a projection of a scene from a
specific viewpoint.

23
Images as Functions

 A gray-tone image is a function:

g(x,y) = val or f(row, col) = val

 A color image is just three functions or a

vector-valued function:

f(row,col) =(r(row,col), g(row,col), b(row,col))

24
Image vs Matrix
Digital images (or just “images”) are typically stored in a matrix.

There are many different file formats.

25
Gray-tone Image as 3D Function

26
Imaging Process
 Light reaches
surfaces in 3D
 Surfaces reflect
 Sensor element
receives light
energy
 Intensity counts
 Angles count
 Material counts

What are radiance and irradiance? 27

Radiometry and Computer Vision*
• Radiometry is a branch of physics that deals with the
measurement of the flow and transfer of radiant energy.

• Radiance is the power of light that is emitted from a

unit surface area into some spatial angle;
the corresponding photometric term is brightness.

• Irradiance is the amount of energy that an image-

capturing device gets per unit of an efficient sensitive
area of the camera. Quantizing it gives image gray tones.

•From Sonka, Hlavac, and Boyle, Image Processing, Analysis, and

Machine Vision, ITP, 1999. 28
CCD type camera:
Commonly used in industrial applications
 Array of small fixed
elements
 Can read faster than
TV rates
 Can add refracting
elements to get
color in 2x2
neighborhoods
 8-bit intensity
common

29
Blooming Problem with Arrays
 Difficult to insulate
adjacent sensing
elements.
 Charge often leaks
from hot cells to
neighbors, making
bright regions larger.

30
8-bit intensity can be clipped
 Dark grid intersections
at left were actually
brightest of scene.
 In A/D conversion the
bright values were
clipped to lower
values.

31
Lens distortion distorts image
 “Barrel distortion” of
rectangular grid is
common for cheap
lenses ($50)
 Precision lenses can
cost $1000 or more.
 Zoom lenses often
show severe
distortion.
32
Resolution
• resolution: precision of the sensor

• nominal resolution: size of a single pixel in scene

coordinates (ie. meters, mm)

• common use of resolution: num_rows X num_cols

(ie. 515 x 480)

• subpixel resolution: measurement that goes into

fractions of nominal resolution

• field of view (FOV): size of the scene a sensor can

33
sense
Resolution Examples
 Resolution
decreases by
one half in
cases at left
 Human faces
can be
recognized at
64 x 64 pixels
per face
34
Image Formats
 Portable gray map (PGM) older form
 GIF was early commercial version
 JPEG (JPG) is modern version
 Many others exist: header plus data
 Do they handle color?
 Do they provide for compression?
 Are there good packages that use them
or at least convert between them?
35
PGM image with ASCII info.
 P2 means
ASCII gray
 Comments
 W=16; H=8
 192 is max
intensity
 Can be made
with editor
 Large images
are usually not
stored as ASCII

36
PBM/PGM/PPM Codes
• P1: ascii binary (PBM)

• P2: ascii grayscale (PGM)

• P3: ascii color (PPM)

• P4: byte binary (PBM)

• P5: byte grayscale (PGM)

• P6: byte color (PPM)

37
JPG current popular form
 Public standard
 Allows for image compression; often 10:1 or
30:1 are easily possible
 8x8 intensity regions are fit with basis of cosines
 Error in cosine fit coded as well
 Parameters then compressed with Huffman coding
 Common for most digital cameras

38
From 3D Scenes to 2D Images
• Object

• World

• Camera

• Real Image

• Pixel Image

39
3D Sensors
 MRA (angiograph)
 Laser range finders showing blood flow.
 CT, MRI, and
ultrasound machines
 Sonar sensors
 Tactile sensors
(pressure arrays)
 Structured light
sensors
 Stereo

40
Where do we go next?

So we’ve got an image, say a single gray-tone image.

What can we do with it?

The simplest types of analysis is binary image analysis.

Convert the gray-tone image to a binary image

(0s and 1s) and perform analysis on the binary image,
with possible reference back to the original gray tones
in a region.

41