0.1

pyMi 0.1 Testing

Thankyou for agreeing to help out with testing with my new app pyMi - written for the Ubuntu App Showdown. This page explains what to do and how to let me know if you have any problems.

You can get in contact with me through Launchpad

For more information on getting the source code and for some screenshots go here http://causticprogramming.blogspot.com.au/2012/06/ubuntu-app-showdown-pymi.html

Before You Begin

Please add my PPA

sudo apt-add-repository ppa:marrabld/stable
sudo apt-get update
sudo apt-get install pymi

The Tests

Please follow through each of these tests and if you find the functionality of the system does not work as expected, file a bug at https://bugs.launchpad.net/pymi/+filebug

For all of the tests below, a new image should be loaded from disk as per the instructions in Test 1

#1 : Load any 8 bit grey scale image

For this test you should be able to load any rasterized 8-bit gray scale image. There are 2 test images in /opt/extras.ubuntu.com/pymi/Images.

  1. Load an image using Menu->Open.

  2. Load an image using the Open icon in the toolbar.
  3. The image should be displayed in the top right window.

Add your name below if this test passed:

* Dan Marrable *

If you find this functionality does not work as expected, file a bug at https://bugs.launchpad.net/pymi/+filebug - please reference the test number and include details of the problem in the bug. Thanks!

#2 : Save image

For this test you should be able to save either of the images as and of the following formats {*.jpg,*.bmp,*.png,*.pdf,*ps,*eps}

  1. Save a file with each of the file extensions above using File->Save menu. The top right images should save.

  2. Save a file with each of the file extensions above using the Save icon. The top right image should save.
  3. Save a file with each of the file extensions above by right clicking on each of the images.

Add your name below if this test passed:

* Dan Marrable

If you find this functionality does not work as expected, file a bug at https://bugs.launchpad.net/pymi/+filebug - please reference the test number and include details of the problem in the bug. Thanks!

#3 : Reduce image resolution

For this test you should be able to reduce the resolution of the image by choosing a particular pixel resolution. (x*y - square only)

  1. Select the Spatial Filter tab and the Resolution Reducer Qtoolbox.
  2. Select a desired (low) resolution and press the Process Image button. The image should appear blurry(ier)
  3. Set the resolution back to the original resolution and press the Process Image button. The Image should go back to the original image. Appear sharp

Add your name below if this test passed:

* Dan Marrable

If you find this functionality does not work as expected, file a bug at https://bugs.launchpad.net/pymi/+filebug - please reference the test number and include details of the problem in the bug. Thanks!

#4 : Grey level transformations

For this test you should be able to perform a grey level transformation by defining different parameters

  1. Select the Spatial Filter tab and the Grey Level Transformation Qtoolbox.
  2. Choose a few different x,y coordinates between 0 and 255 separated by commas. e.g 127,255,255,0
  3. Press the Process Image button.
  4. The top image should transform and the bottom image should plot the transformation.
  5. Chose the Log Transformation radio button and choose an 'a' value for the function.
  6. The top image should transform and the bottom image should plot the transformation.

Add your name below if this test passed:

* Dan Marrable

If you find this functionality does not work as expected, file a bug at https://bugs.launchpad.net/pymi/+filebug - please reference the test number and include details of the problem in the bug. Thanks!

#5 : Histogram equalization

For this test you should be perform histogram equalization according to http://en.wikipedia.org/wiki/Histogram_equalization

  1. Select the Spatial Filter tab and the Histogram equalization Qtoolbox.
  2. Chose the Plot Histogram radio button and press the Process Image button.
  3. The bottom image should plot the histogram of the top image.
  4. Chose the Histogram Equalization radio button and the Process Image button.
  5. Two figures should appear. Figure 1 should show the before and after histograms. Figure 2 should show the cumulative distribution function.
  6. After closing the figures, the top image should be the equalized image and the bottom image should show the histogram of the image.

Add your name below if this test passed:

* Dan Marrable

If you find this functionality does not work as expected, file a bug at https://bugs.launchpad.net/pymi/+filebug - please reference the test number and include details of the problem in the bug. Thanks!

#6 : Moving average filters

For this test you should be able to perform a low pass filter in the spatial domain using 'box car' and 'median average' filters.

  1. Select the Spatial Filter tab and the Moving Average Filters Qtoolbox.
  2. Chose the Box Car Filter radio button and the size of the filter in x,y pixels and press the Process Image button
  3. The top right image should appear more blurry with increasing box car size.
  4. Repeat steps 2 and 3 for the Median Filter radio button.

Add your name below if this test passed:

* Dan Marrable

If you find this functionality does not work as expected, file a bug at https://bugs.launchpad.net/pymi/+filebug - please reference the test number and include details of the problem in the bug. Thanks!

#7 : Convolution filters

For this test you should be able to define your own square kernel to convolve with the image.

  1. Select the Spatial Filter tab and the Convolution Filters Qtoolbox.
  2. Define a kernel to convolve in the kernel text box eg [1,1,1,1,4,1,1,1,1] will be [1,1,1];[1,4,1];[1,1,1]/<sum of the weights>.

  3. Press the Process Image button
  4. The image should be altered according to the kernel.
  5. Some examples can be found at the bottom of this page http://docs.gimp.org/en/plug-in-convmatrix.html

Add your name below if this test passed:

* Dan Marrable

If you find this functionality does not work as expected, file a bug at https://bugs.launchpad.net/pymi/+filebug - please reference the test number and include details of the problem in the bug. Thanks!

#8 : Edge detection

For this test you should be able to run an edge detection kernel over the image.

  1. Select the Edge Detection tab
  2. Make sure the Re-scale image check box is checked. This should re-scale the image if their are negative edges when ticked the negative edges should appear black and the positive edges white - comparability. If un-ticked, no negative edges should show.

#8.1 Prewit

  1. Select the Prewitt qtoolbox and the Horizontal Prewitt radio button.
  2. Press the Process Image button.
  3. The image should accentuate the horizontal lines in the image.
  4. Repeat steps 2 and 3 for the Vertical Prewitt radio button

#8.2 Sobel

  1. Select the Sobel qtoolbox and the Horizontal Sobel radio button.
  2. Press the Process Image button.
  3. The image should accentuate the horizontal lines in the image, more noticeably that the Prewitt - depending on the image.
  4. Repeat steps 2 and 3 for the Vertical Sobel radio button

#8.3 Roberts

  1. Select the Roberts qtoolbox and the top radio button.
  2. Press the Process Image button.
  3. The image should accentuate the diagonal lines in the image - depending on the image.
  4. Repeat steps 2 and 3 for the second radio button

#8.4 Laplace

  1. Select the Laplace qtoolbox and the radio button.
  2. Press the Process Image button.
  3. The image should accentuate all of the lines in the image, more noticeably that the Prewitt - depending on the image.

#8.5 Unsharp Mask

  1. Select the Unsharp mask qtoolbox
  2. Choose the size of the low-pass filter to remove from the image in x*y pixels.
  3. Press the Process Image button.
  4. The image should appear sharper - if the image is already sharp, try performing a box car filter on the image first

Add your name below if this test passed:

*

If you find this functionality does not work as expected, file a bug at https://bugs.launchpad.net/pymi/+filebug - please reference the test number and include details of the problem in the bug. Thanks!

#9 : Noise

For this test you should be able to add varying degrees of noise to an image.

  1. Select the Noise tab and the the Random Noise radio button.
  2. Choose a magnitude of noise.
  3. Press the Process Image button.
  4. The image should appear to have 'white noise' over it.
  5. Higher magnitudes should make the image appear more noisy.
  6. Repeat steps 2 and 3 for Salt and Pepper
  7. The image should appear to have 'spotty noise'

Add your name below if this test passed:

*

If you find this functionality does not work as expected, file a bug at https://bugs.launchpad.net/pymi/+filebug - please reference the test number and include details of the problem in the bug. Thanks!

#10 : Fourier Analysis

For this test you should be able to perform a Fourier analysis on an image.

  1. Select the Fourier tab and the Fourier Transform radio button.
  2. Press the Process Image button.
  3. The bottom image should show the Discrete Fourier Transform of the image above.
  4. Chose the Inverse Fourier Transform radio button and click the Process Image button.
  5. Repeat step 1 and 2 with the Log Transform check box ticked.
  6. The bottom image should have a brighter perimeter.
  7. Inverse Fourier Transform should not work if Log Transform is checked.

Add your name below if this test passed:

*

If you find this functionality does not work as expected, file a bug at https://bugs.launchpad.net/pymi/+filebug - please reference the test number and include details of the problem in the bug. Thanks!

#11 : 2D Fourier Mask

For this test you should be able to perform low-pass filters by applying a mask in the Fourier domain.

  1. Chose the Fourier tab and the 2D Fourier Mask qtoolbox.
  2. Chose the Gaussian filter radio button.
  3. Choose a standard deviation and press the Process Image button.
  4. The image should appear more blurry with lower standard deviations.
  5. The bottom image should display the mask.
  6. Choose the Butterworth radio button.
  7. Vary the Cuttoff frequency and a filter order and press the Process Image button.
  8. Lower Cuttoff frequencies should make the image blurrier and higher filter order should make the cutoff steeper.
  9. The bottom image should display the mask.

Add your name below if this test passed:

*

If you find this functionality does not work as expected, file a bug at https://bugs.launchpad.net/pymi/+filebug - please reference the test number and include details of the problem in the bug. Thanks!

#12 : Signal Generator

In this test you should be able to add signals together and see the Fourier representations in real-time

  1. Choose the Signal Generator tab.
  2. Adjust and vary the sliders to add signals together of varying frequencies and amplitudes.
  3. The top image should show the signals in the spatial domain and the bottom image should show the signals in the frequency domains.

Add your name below if this test passed:

*

If you find this functionality does not work as expected, file a bug at https://bugs.launchpad.net/pymi/+filebug - please reference the test number and include details of the problem in the bug. Thanks!

#13 : 2D Signal Generator

In this test you should be able to add 2D signals together and see the Fourier representations in real-time

  1. Choose the 2D Signal Generator tab.
  2. Adjust and vary the sliders to add signals together of varying frequencies and amplitudes.
  3. The top image should show the signals in the spatial domain and the bottom image should show the signals in the frequency domains.

Add your name below if this test passed:

*

If you find this functionality does not work as expected, file a bug at https://bugs.launchpad.net/pymi/+filebug - please reference the test number and include details of the problem in the bug. Thanks!

Pymi/GetInvolved/Testing/0.1 (last edited 2012-07-06 08:12:28 by 58-7-197-117)