innfiniteminds/antarctic_image_processing

Frame by frame analysis of a NASA video to collect information on the extent of sea ice in the Antarctic. Project for ELEC 221: Signals and Systems (Sep-Dec 2016).

ELEC 221 2016W1 Project: Climate Change and The Antarctic Sea Ice Extent - Visual to Statistics

MATLAB code from my ELEC 221 Project, completed back in 2016.

ELEC 221: 2016W1

ELEC 221, Signals and Systems at UBC covers the following topics :

Complex numbers, LTI systems, convolution sum, discrete-time Fourier series and transforms, z-transform, sampling, introduction to filtering and modulation, feedback systems, stability.

The course had an optional project to be implemented in MATLAB, worth upto 10% of our final course grade. The project specification, report draft, and source code are available in this repository.

Notes

The code was slightly modified and refactored from the original version. The entire script has been separated into different files for modularity and easier disk management.

Instructions

1. Clone this repository

2. Open MATLAB, and navigate to this directory on your local machine

3. Run the script titled "ExtractFrames.m". After completion of this script, all 3081 frames should be extracted and stored in the directory.

4. Run the script titled "DeleteExcessOGFrames.m". Since we are interested in only the frames 1359-1699, we will remove the rest to free up storage.

5. Run the script titled "ConvertToGrayScale.m". This will make a new GrayScale copy of the images.

6. Run the script titled "DeleteRemainingOGFrames.m". This will delete the original frames of interest as we no longer need them.

7. Run the script titled "MaskAndConvertBW.m". This will make Black and White copies of the GrayScale images, and then make another copy of each where the date is masked out.

8. Run the script titled "DeleteBWFrames.m". Since we now have the masked images, we can delete the raw black and white ones.

9. Run the script titled "Filter.m". This will use a Gaussian Low-Pass Filter to remove any effects from shadows.

10. Run the script titled "DeleteMasked.m". This will delete all of the masked images, since we will work with the filtered ones now.

11. Run the script titled "IceChangePlot.m". Observe the plots.

12. Finally, run the script "DeleteFilteredFrames.m". This will delete all the filtered images from the directory.