0.1
Assignment 3 Answers
0.2
Answer: 1
1. A image denoising technique will work here which can remove the isolated dots. We can perform an adaptive weighted mean filtering algorithm.
2. Lack of sharpn
1
Lecture 4
Image Enhancement II
2
Spatial Filtering
Filter
In general, a tool to suppress or emphasize some aspect of
signal
Accepting or rejecting certain frequency components
E.g. lowpass filte
Digital Image Processing, 3rd ed.
Matrices and Vectors
Gonzalez &
Woods
www.ImageProcessingPlace.com
Lectures 2
REVIEWS
Matrices
Probability and Statistics
Linear Systems
19922008 R. C. Gonzalez & R.
1
Lecture 3
Image Enhancement
2
Image Enhancement: improving the quality of images
(make it better) Or modifying image to bring out hidden
features. Two kinds: spatial domain and frequency
domain.
Ima
Concordia University
Department of Computer Science
& Software Engineering
COMP 478/6771 Image Processing
Assignment 1
Due Date: September 29, 2014
Part I: Theoretical questions
1. Consider two 8-bit
Concordia University
Department of Computer Science
& Software Engineering
COMP 478/6771 Image Processing
Solutions to Assignment 1
Question 1.
(a) Pixels are integer values, and 8 bits allow represen
Concordia University
Department of Computer Science & Software Engineering
COMP 478/6771 Image Processing
Assignment 3
Due date: November 11, 2014
Theoretical questions
1. Let I A , I B be two complet
Concordia University
Department of Computer Science & Software Engineering
COMP 478/6771 Image Processing
Solutions to Assignment 2
Part I: Theoretical questions
Problem 1.
Problem 2.
Problem 3.
Probl
Discrete Fourier Transform
Javier Montoya
Photogrammetry and Remote Sensing
ETH Zurich
March 16, 2012
1
Introduction
The Discrete form of the Fourier transform is known as Discrete Fourier Transform (
Image Processing
Transforms of 2D signals
Instructor: T. D. Bui
Email: [email protected]
Department of Computer Science
and Software Engineering
Concordia University
1
Differences Between 1D and Mu
Filtering in Frequency
Domain
Instructor: T. D. Bui
Email: [email protected]
Department of Computer Science
and Software Engineering
Concordia University
1
Image processing in the frequency domain
Ans1: Lets suppose we have an image f(x,y) with value 1 at its center of
size and every where zero. We take N as 64.
A filter h(x,y) with a mask of 3 3, and each pixel has a value of 1/9.
Now, filteri
0.1
Assignment 4 Answers
0.1.1
Part(a)
Repeatedly eroding an image will lead to a single pixel element.
For the dilated image it will grow to infinity
0.1.2
Part(b)
The smallest image will be the stru
1
Canny edge detector
A computational Approach to Edge Detection,
J. Canny, IEEE PAMI, no.6, pp 679-698,1986
More complex but has better performanc e than all methods
discussed so far.
Based on firs
Ans: Considering both triangles to be congruent, height for the flat area will be:
24/50=h/.75
H=.36m or 360mm
Sensor has a resolution of 24/3712=.0065mm in the horizontal direction
So same resolution
Topics in Image Processing
Point, Line, Edge Detection, and Image
Segmentation
Instructor: T. D. Bui
Email: [email protected]
Department of Computer Science
and Software Engineering
Concordia Unive
1
Morphological Processing
2
Operations on Sets of Points
3
Other Set Operations
4
Dilation
reflecting about origin
5
Dilation Example
B
A
A
B
The dilation of A by B is the set of all displacements
su
Filtering in Frequency
Domain
Instructor: T. D. Bui
Email: [email protected]
Department of Computer Science
and Software Engineering
Concordia University
1
Image processing in the frequency domain
COMP 478/6771
Image Processing
Introduction
Instructor: T. D. Bui
Email: [email protected]
Department of Computer Science
and Software Engineering
Concordia University
1
COMP 478/6771
Instructor:
O
Concordia University
Department of Computer Science
& Software Engineering
COMP 478/6771 Image Processing
Assignment 2
Due Date: October 20, 2014
Theoretical questions
1. Do problem 3.20 of the 3rd Ed