|
Lana Carnel
Learning Dynamics and Tracking
  
  
Image Segmentation Using HSV Colorspace
|
Weekly Updates
|
Week 1
Background research on previous segmentation and camera network explorations. Write preliminary paper using this background information with respect to a new possible application.
|
Week 2
Refresh knowledge of Matlab and obtain knowledge of image processing
toolbox. Analysis of previously written codes using images in an array.
Obtain video clips of desired camera network by simulating tracking
situation with color targets placed on remote controlled cars. |
Week 3
Write two function files to perform color segmentation on established target color. Initial function using RGB color space, the second using HSV. Examine RGB vs. HSV color space. Found HSV to reduce input segmentation parameters from three to 1 (in addition to a tolerance variance) by using the intensity layer. Image source and constructed files are below.
|
Week 4
Select pertinent clips from video obtained in week 2. Adapt function
files to perform written segmentation functions to simulation video.
Add and revise preliminary paper to reflect completed work. |
Week 5
Optimize function files to the image array. Expand HSV function to perform
segmentation using a combination of all three layers of the colorspace.
Write function to find centroid of individual image. Adapt centroid
function to the array of images. Adapt centroid function to cluster
multiple targets in a single frame.
Week 6
Optimize files for minimum run time. Segmentation and identification
requires approximately 2 seconds per frame.
Week 7
Power point presentation to CHESS/SUPERB group. Work on additional
methods of segmentation. Begin work on final deliverables.
Week 8
Construct and print poster for final presentation.Write, revise, edit
final paper. Final poster presentation. |
