CV2007_1_HW05
20 points total

Use the functions that you wrote for HW#4.

Part A: (15 points) Write a Matlab function that will accomplish the following:
Your function must have the following minimum functionality to obtain the 10 points.
1. Name your function CV2007_1HW05your_name and replace your name with your actual name.
2. The function should take as the input parameter the name of an image file,
for example: CV2007_1HW05RogerGaborski('OrangeBuilding.jpg'). There is only one input parameter
3. Display the image
4. Display the message: 'Click on sample points. Double click on last point'
5. Use impixel to collect the data.
6. Request a threshold value from the user and display: 'Enter threshold value between 1 and 255  '
7. Use the user supplied threshold to segment the image - DO NOT PASS THE THRESHOLD VALUE AS A FUNCTION PARAMETER VALUE
8. Use your the basic code from your colorsegment (from Part A), but this time the data points and threshold value supplied by the user to segment the image - use both the Euclidean and Mahalanobis  methods
9. Display and label the Euclidean method of color segmentation
10. Display and label the Mahalanobis method of color segmentation
11. Print to the screen the average color density values and the covariance matrix.

Part B: (5 points) Write a description of how the algorithm implemented in colorsegment operates. Specifically describe the difference between the Euclidean and Mahalanobis methods.
Process the image OrangeBuilding (on webpage) as described below:
    -Result 1 OrangeBuilding: select points to segment the orange building using both methods. Describe results and justify your results - why did you get your results? Do not use more than 10 sample points.
    -Result 2 OrangeBuilding: select points to segment the sky using both methods. Describe results and justify your results - why did you get your results? Do not use more than 50 sample points.
For each image include the average color values and covariance matrix.
   
e-mail copies of the images (total 4), code, algorithm description and writeup.