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 sharpness can be improved by adding high frequency components
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 filter, highpass filter, bandpass filter
Lowpass filter: sm
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. E. Woods
Digital Image Processing, 3rd ed.
Matrices an
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.
Image Restoration: correcting images subjected to
noise, b
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 images whose intensity levels span the full range from
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 representation of 256 contiguous integer values. In our
work, t
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 completely different gray level images, and FA , FB be
the Fou
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.
Problem 4.
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 (DF T ) and in 1D is given by
the following equation:
N
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 Multi-Dimensional
Signal Processing
Much more data for M
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 relies on the
Fourier Transforms.
A good understandin
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, filtering operation is done by convolving h(x,y) with f(x,y) a
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 structuring element i.e. the single element.
For the dilate
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 first order derivative s, and directions .
Smoothed image
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 will be kept for the flat area. In 1mm camera can reso
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 University
1
Image segmentation:
- computer vision & image u
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
such that they overlap by at least one element.
6
Another
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 relies on the
Fourier Transforms.
A good understandin
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 Edition of Gonzalez and Woods.
2. Do problem 4.1 of the 3