From:                              michaelpayne@gmail.com on behalf of Mike Payne <paynemi@auburn.edu>

Sent:                               Sunday, February 26, 2012 10:11 PM

To:                                   Thaddeus Roppel

Subject:                          Re: HW4

 

Of course, sorry I didn't include one in the original message.

I loaded the images into MATLAB and converted them to grayscale, which basically just gives you an intensity for every coordinate (pixel) in the image.  Then I normalized those values by converting the original uint8 image into double.  This new image was then filtered: every cell with a value less than 0.55 was rewritten to zero, and all others were left alone.  The resulting image was then used to create a mesh image and that image was viewed from beneath, as for some reason the image got flipped during the process.

Here's the exact code I used:

p1 = rgb2gray(imread('p1.jpg'));
p1a = im2double(p1);
ix = 1;
iy = 1;
for ix = 1:1200
    for iy = 1:1600
        if p1a(ix,iy) < 0.55
            p1b(ix,iy) = 0;
        else
            p1b(ix,iy) = p1a(ix,iy);
        end
    end
end
mesh(p1b);
view(0,-90);

On Sun, Feb 26, 2012 at 9:11 PM, Thaddeus Roppel <ROPPETH@auburn.edu> wrote:

Got it. Can you also provide a brief writeup that explains how you got these images?

 

From: michaelpayne@gmail.com [mailto:michaelpayne@gmail.com] On Behalf Of Mike Payne
Sent: Sunday, February 26, 2012 7:34 PM
To: Thaddeus Roppel
Subject: HW4

 

Dr. Roppel,

Attached is my solution for homework 4.  I've got an easily-identifiable path in all of the images, but I don't know how to go about drawing a line on those images to show the robot's path -  just don't have the necessary experience with image manipulation and sorting algorithms in MATLAB.


--
Mike Payne
Graduate Research Assistant
Auburn University
paynemi@auburn.edu
205-903-9586




--
Mike Payne
Graduate Research Assistant
Auburn University
paynemi@auburn.edu
205-903-9586