Fix get_mask and display_roi

[VenissaCarolQuadros]

1. get_mask: Fixing 'valueerror: too many values to unpack (expected 2)' in case of rgb (or >2 dimension) images.
2. display_roi: Quick fix . Discretion advised during merge... 'display_roi' was not displaying image earlier

[jdoepfert]

tx for the contribution! If you have RGB arrays, the function get_mean_and_std() will be undefined though..
So support for RGB will be more complex to handle, it needs more than a quick fix. One option would be to only allow get_mean_and_std for 2D arrays.

Also: Pls do submit one pull request at a time for each change (i.e. not putting the display_roi quick fix into the same PR as the valueerror fix).

[jdoepfert]

Thinking about it, rather than fixing the code, you should rather call get_mask() like so:

roi.get_mask(self, img[:,:,0])

[jdoepfert]

I will update the documentation to make this clearer

[jdoepfert]

Can you tell me what was the issue with display_roi and how it can be reproduced?

[VenissaCarolQuadros]

tx for the contribution! If you have RGB arrays, the function get_mean_and_std() will be undefined though..
So support for RGB will be more complex to handle, it needs more than a quick fix. One option would be to only allow get_mean_and_std for 2D arrays.

Also: Pls do submit one pull request at a time for each change (i.e. not putting the display_roi quick fix into the same PR as the valueerror fix).

Ahh yes! Sorry for not taking that into account. Do not want to complicate this further however I could suggest going with mask=mask if image.ndim!=3 else np.dstack(tuple([mask for i in range(image.shape[2])])) for the sole reason that the mean and std would yield different results depending on whether you chose img[:, :, 0] or img[:, :, 1]. Do you see any problems that could arise with this ?

[VenissaCarolQuadros]

Can you tell me what was the issue with display_roi and how it can be reproduced?

The issue I was facing with display_roi is that it refused to display the image.. I followed the instructions as in the documentation with plt.imshow(image) ; my_roi = RoiPoly(color='r') ; plt.imshow(image) ; my_roi.display_roi() but the image window fails to pop up.

I use Anaconda Command Prompt ...

No errors show up whatsoever hinting that perhaps plt.show() could help..

P.S: Is this an eccentric issue that's affecting only me or is there something I'm missing? It could be a problem with my environment then... :(