Computer Science - The City College of New York
CSC I6716 - Spring 2014 3D Computer Vision
Assignment 2. Feature Detection ( Due: March 11th before
class)
Note: Turn in a document (in
writing; please type) containing a list of your .m files,
images showing the results of your experiments, and an analysis of
the results.All the writings must be hard copies in print - you may
hand them in during class, or put them in my mailbox in the CS main
office . You also need to turn in your “soft” copies of your
assignment via email. For the programming part, send ONLY your source code to me by
email – please don’t send in your images and
executable (even if you use C++). You are responsible for the
lose of your submissions if you don’t write “CSC I6716 Computer Vision Assignment 2”
in the subject of your email. Do write your names and IDs (last four
digits) in both your hard copy and soft copy submissions.
Choose one or two of the images on the course web page for this
assignment or use an image of your own choice. If you use a
different image, be careful that the image has not been saved under
JPEG using a high level of compression. This often introduces
artifacts into the image that confound subsequent analyses of the
results. For the most part, you should apply your operators to
a gray-scale version of the image.
1. (20 points) Generate the histogram of the image you are
using . If it is a color image, please first turn it
into an intensity image and then generate its histogram. Try
to display your histogram, and make some observations of the image
based on its histogram.
2. (20 points) Apply the 1x2 operator and Sobel operator to your
image and analyze the results of the gradient magnitude
images. Does the Sobel operator have any clear visual
advantages over the 1x2 operator? If you subtract the 1x2 edge
image from the Sobel are there any residuals? (Note: don't forget to normalize your
results as shown in slide # 29 of feature extraction
lecture: part 2)
3. (20 points) Generate edge maps of the above gradient maps (20
points). You may first generate a histogram of each gradient
map, and only keep certain percentage of pixels
(e.g. 5% of the highest gradient values) as edge pixels
(edgels) . Use the percentage to find a threshold for the gradient
magnitudes.
4. (20 points) What happens when you increase the size of the kernel
to 5x5 , or 7x7? Discuss computational cost (in terms of members of
operations, and the real machine running times), edge detection
results and sensitivity to noise, etc. Note that your larger kernel
should still be an edge detector.
5. (20 points) Suppose you apply the Sobel operator to each of
the RGB color planes comprising the image. How might you
combine these results into a color edge detector? Do the
resulting edge differ from the gray scale results? How and
why?
6. Proof-reading (10 bonus points). While you are reading the lecture
notes on feature extraction, please write down your
suggestions of changes for typos, unclear sentences, and etc.., in a
separate file and send it to me via email. Please do not directly
work on the PDF file. Instead please provide a list of your changes,
including page numbers, the original and the changed parts. You
might also put your comments for me to further improve the writing
of the document.