2011-08-30 E0005E Industrial Image Analysis 1 › cms_fs › 1.36192! › file › E0005E_Lecture1_Digital… · - login and go to the course room E0005E Industrial Image Analysis

2011-08-30 E0005E Industrial Image Analysis 1

INSTITUTIONEN FÖR SYSTEMTEKNIK

LULEÅ TEKNISKA UNIVERSITET

2011-08-30 E0005E, Course Introduction 2

E0005E Industrial Image Analysis

Matthew Thurley, Teacher

Anders Landstöm, Lab Assistant

[email protected]

[email protected]

mailto:[email protected]

mailto:[email protected]

What you need to know

• Course Study guide (read this carefully)

- Course structure, topics, assessment criteria

- This course has a problem based learning approach

- The assignments are Development Phases in the production of an industrial measurement system

• Course web page http://www.ltu.se/edu/course/E00/E0005E/E0005E-Industriell-bildanalys-1.36192?l=en

• Fronter https://fronter.com/ltu/

- Has a virtual room with general course information.

- Submit assignments in Fronter

- login and go to the course room E0005E Industrial Image Analysis H1* There is a Submissions link in the left column.

- Submit in the correct folder. If you are late you will have to submit in the Late Submissions folder


http://www.ltu.se/edu/course/E00/E0005E/E0005E-Industriell-bildanalys-1.36192?l=en







https://fronter.com/ltu/

What you need to know

• Attendance – Make sure you are registered

• Plagarism - ”copying another persons work and presenting it as your own”

- Watch this video

- http://www.youtube.com/watch?v=Mwbw9KF-ACY&feature=related

- Press the CC button to turn english subtitles on or off

- Ephorus compares your submission against millions of documents from the internet and against past submissions in this course

2011-08-30 E0005E, Lecture 3 4

http://www.youtube.com/watch?v=Mwbw9KF-ACY&feature=related





Ephorus Example 4% Match



Labs

• Labs must be done individually (you can of course discuss problems and solutions with your friends but you must do your own coding and report writing)

• Lab 8 may have some group work, but you still must write your own report

• Helpful documents

- MATLAB-Tips.pdf

- ReportTemplate.doc


The lab environment

• A2506 in the A-building

• We will use Matlab for the assignments

Variable and

directory

browser

Command

history

Command window

DIGITAL IMAGES

Matthew Thurley

E0005E Industrial Image Analysis

2011-08-30 E0005E, Industrial Image Analysis 8


Light

Imaging based on Gamma Rays • Nuclear medicine,

Injecting radioactive isotopes into the body that emit gamma rays (a) or cause the emission of gamma rays (b)

• (c) Gamma ray band image of the ”Cynus Loop” interstellar gas cloud, the remnant of an exploded star

• (d) Gamma ray image of a valve in a nuclear reactor


Imaging based on Gamma Rays

“An image taken by a gamma ray camera showing the bottom of a ventilation stack standing between Fukushima Daiichi nuclear power plant's No.1 and No.2 reactors, where radiation exceeding 10 sieverts per hour - seen here in red” 2011-08-02 blog post

“a level that could lead to incapacitation or death after just several seconds of exposure”

http://photoblog.msnbc.msn.com/_news/2011/08/02/7227954-lethal-levels-of-radiation-recorded-in-fukushima-gamma-ray-image


























X-ray Imaging • Medicine, familiar chest x-ray (a). An x-

ray source behind the patient sends x-rays through the patient towards a detector on the other side. X-rays are absorbed by tissue and bone reducing the x-ray strength arriving at the detector creating different intensities in the image

• (b) An xray contrast medium has been injected into the aorta of the patient to highlight parts of interest

• (c) X-ray CAT scan (computerized axial tomography)

• (d) Industrial X-ray imaging for defect detection

• (e) Cynus Loop dust cloud

2011-08-30 12

Ultraviolet Imaging • Flourescence microscopy is a field of

miscroscope imaging where ultraviolet light is used to excite the electrons in a material causing them to flouresce (emit light). Some material flouresce naturally and flourescent chemicals can be applied to a sample

• (a) Miscroscope image of a healthy corn sample treated with flourescent chemicals

• (b) Corn infected by ”smut” disease

• (c) Cynus Loop dust cloud in the high energy ultraviolet band

2011-08-30 13

• Remote sensing

Visible & Infrared Imaging

2011-08-30 14

• Industrial Imaging

• Automated visual inspection of manufactured goods

Visible & Infrared Imaging

2011-08-30 15

• Most common application is radar

• Some radar waves can penetrate clouds and even see through ice (martian polar ice), vegetation or dry sand (sahara desert).

• Imaging radar is an active illumination technique (like a camera with a flash)

• Spaceborne radar can see through clouds and take hi resolution images of the land.

• Consider NASAs detailed radar imaging of Venus which is permanently shrounded in sulfuric acid clouds

Imaging in the Microwave Band

2011-08-30 16

Image courtesy of JPL

• Magnetic resonance imaging - Pulsed radio waves pass through the patient, radio pulses emminate from the patients body. Complex algorithms determine the source location and strength of these emminations reconstructing a picture of the patients body part.

• Crab Pulsar – naturally emitting source of radio waves

Imaging in the Radio Band

2011-08-30 17

• Sound

• Ultrasound (medical imaging)

• Siesmic imaging (oil and gas exploration)

• Electron miscroscopy

Imaging from other sources

2011-08-30 18


Visible Light

• Visible Light is the part of the electromagnetic spectrum that causes a reaction in our visual systems

• Generally these are wavelengths in the range of about 350-750 nm (nanometers)

• Long wavelengths appear as reds and short wavelengths as blues

Three-Color Theory

Human visual system has two types of sensors

• Rods: grayscale, highly sensitive to intensity, superior at night & peripheral vision

• Cones

- Color sensitive

- Three types of cone

- Only three values (the tristimulus values) are sent to the brain

Need only match these three values

• Need only three primary colors

Speaking Notes

• Initial processing of light in most optical systems is based on the human visual system, however, the human visual system has a back end far more complex than any camera.

• The optic nerves are connected to the rods and cones in an extremely complex arrangement with many of the characteristics of a sophisticated signal processor.

• The sensors in the eye do not react uniformly to light energy of different wavelengths (colors)

• Define intensity as the physical measure of light energy

• Define brightness as the measure of how intense we percieve the light to be.

2011-08-30 20

Color Perception http://en.wikipedia.org/wiki/Color_perception

• Cone cells in the human eye.

• We are most sensitive to green light where all the ranges overlap

• However, red light stimulates almost exclusively L-cones, and blue light almost exclusively S-cones.

Cone

Type

Range

(nm)

Peak

Wavelength

S (blue) 400..500 420-440 nm

M (green) 450..630 534-545 nm

L (red) 500..700 564-580 nm


Image: WikiCommons

2011-08-30 E0005E, Digital images 22

Digital recording medium

• CCD Charged-Coupled Device

• Bayer mask


Color and monochromatic representation

• Chromatic light (color)

• Monochromatic light

Prism White light

Intensity of red light

Intensity of green light

Intensity of blue light

White light Intensity of white

light (or grey level)

What is a photographic digital image

• Grid of elements (pixels)

• Each element has a color

• Each color is represented by a number

What is a photographic digital image

• Grid of elements (pixels)

• The image (grid) has a number of rows, and a number of columns

• We identify each pixel by specifying its position (i,j) in the image using a row number i, and a column number j

column j

row i


Digital image acquisition process


Image sampling and quantization


Image sampling and quantization


Spatial Resolution (Sampling)

• Spatial resolution is a measure of the smallest perceptible detail in an image

- Dots per unit distance

- Dots per inch (DPI)

• 20-megapixel camera can be expected to have better imaging capabilities than a 8-megapixel camera.

• But, size of image does not tell complete story.


Spatial Resolution


Intensity Resolution (Quantization)

• Intensity resolution is a measure of the smallest perceptible change in intensity level in an image

- Integer power of two

- 8 bits often used (256 gray levels)

• ‘Smallest perceptible change’

- Not how the human percieve

- Statement on how the intensity is quantized.


Image sampling and Quantization


Intensity Resolution (Quantization)


A few operations

• We will go through

- Point operations

- Histogram calculation

- Contrast stretching

- Thresholdning


Point operations

• A point operation is an operation where the gray level gi at a certain pixel (i,j) is replaced with a new gray level go according to some mapping go = F(gi)

(i,j)

gi = I(i,j)

go = F(gi)

I(i,j) = go


Point operations

• Point operations is implemented as a Look Up Table

- If the mapping F(g) is independent of pixel position

2 9 0 1

0 7 4 3

9 8 1 6

3 0 6 2

Original image New image

Look Up Table

0 2 4 8 16 16 16 16 32 64

0 1 2 3 4 5 6 7 8 9


Histogram calculation

• The histogram of a digital image with intensity levels in the range [0,L-1] is a discrete function

h(rk) = nk

rk - is the kth intensity value

nk - is the number of pixels in the image with intensity rk

• Commonly normalized

p(rk) = nk / NM






Contrast stretching

• Piecewise-Linear Transformation

- Can design complex transformation functions

• The shape of the transformation

- Controlled by (r1,s1) and (r2,s2)

• Typical transformation for

contrast stretching


Contrast stretching, an example


Contrast stretching, an example


Thresholding

• Creates a binary image

• Useful for identifying objects in images - Calculate size, shape etc.

• Study the histogram to find appropriate threshold values

Reading

• Digital Image Processing (Third Edition) Gonzalez and Woods

- Ch 1

- Ch2 2.1-2.5


Summing Up

• Consider the following three questions;

- What do I need to work on?

- What have I learnt today?

- What was the main point left unanswered today?

• Write your answers in the provided journal. Write the lecture number 1 on top of the page. Write your name and student number on the back of the book


2011-08-30 E0005E Industrial Image Analysis 1 › cms_fs › 1.36192! › file › E0005E_Lecture1_Digital… · - login and go to the course room E0005E Industrial Image Analysis

Documents