<div dir="ltr">See <a href="http://scikit-image.org/docs/dev/user_guide/numpy_images.html">http://scikit-image.org/docs/dev/user_guide/numpy_images.html</a> and the discussion at <a href="https://github.com/scikit-image/scikit-image/pull/1280">https://github.com/scikit-image/scikit-image/pull/1280</a>.<div><br></div><div>skimage documents [rr, cc, ch] as the 'standard' for RGB(A) images (<a href="http://scikit-image.org/docs/dev/user_guide/numpy_images.html#coordinate-conventions">http://scikit-image.org/docs/dev/user_guide/numpy_images.html#coordinate-conventions</a>)<br><div><br></div><div>In general RGB color channels have a very different meaning that planes. The former is a result of internal details of a camera, where as the later is the result of changing something in data acquisition (in cases where color is one of the experimental controls it is a plane because you are probably not collecting exactly RGB and are going to composite down to get a RGB (false) color image).</div><div><br></div><div>Tom </div></div></div><br><div class="gmail_quote"><div dir="ltr">On Wed, Aug 5, 2015 at 9:09 AM Benjamin Root <<a href="mailto:ben.root@ou.edu">ben.root@ou.edu</a>> wrote:<br></div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div dir="ltr"><div>I think it has to do with the conventions used by PIL way back in the day. To me, the current conventions make sense to me because imshow can work with just a simple 2D image. Color is then additional dimensions (and thus at the end). If I slice up an image like so: `im[20:40, 50:80]`, then imshow will work as expected regardless if that image was a 2D grayscale image or a 3D RGB[A] image.<br><br></div><div>Cheers!<br></div>Ben Root<br></div><div class="gmail_extra"><br><div class="gmail_quote">On Wed, Aug 5, 2015 at 4:02 AM, Fabien <span dir="ltr"><<a href="mailto:fabien.maussion@gmail.com" target="_blank">fabien.maussion@gmail.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">Hi all,<br>
<br>
just to follow up a quite intense discussion on the numpy mailing list about dimensions ordering: why are matplotlib RGB images of dimensions (row, column, plane) instead of the standard (plane, row, column)?<br>
<br>
I guess this has been asked thousand times but I can't seem to find the answer...<br>
<br>
Thanks!<br>
<br>
Fabien<br>
<br>
_______________________________________________<br>
Matplotlib-users mailing list<br>
<a href="mailto:Matplotlib-users@python.org" target="_blank">Matplotlib-users@python.org</a><br>
<a href="https://mail.python.org/mailman/listinfo/matplotlib-users" rel="noreferrer" target="_blank">https://mail.python.org/mailman/listinfo/matplotlib-users</a><br>
</blockquote></div><br></div>
_______________________________________________<br>
Matplotlib-users mailing list<br>
<a href="mailto:Matplotlib-users@python.org" target="_blank">Matplotlib-users@python.org</a><br>
<a href="https://mail.python.org/mailman/listinfo/matplotlib-users" rel="noreferrer" target="_blank">https://mail.python.org/mailman/listinfo/matplotlib-users</a><br>
</blockquote></div>