Mailman 3 October 2020 - scikit-image

Big image sprint at SciPy Japan
by Genevieve Buckley 31 Oct '20

31 Oct '20

Juan and I are going to run a "big image" sprint at SciPy Japan next week (Nov 1-2, 9am-5pm Tokyo time). Anyone who wants to join the sprint (even asynchronously) is welcome to join! There's a hackmd here with some more details about topics people might like to sprint on: https://hackmd.io/qpiRFdIsTdCAfdr_C2mMUA - Genevieve

2 1

Developers' meeting on Thursday October 29th at 11am CEST
by Marianne Corvellec 27 Oct '20

27 Oct '20

Dear all, The next developers' meeting for scikit-image will take place next Thursday (October 29th) at 11am CEST (Paris time). All developers are welcome to attend the meeting. This meeting will focus again on the preparation of the 0.18 release. Other agenda items can be found at: https://hackmd.io/r5oMbIyvQH2K0mhgntAW2g We will use Zoom (https://monash.zoom.us/j/284282585). Best, Marianne

1 1

multidimensional (4D) color conversion
by Matteo Niccoli 17 Oct '20

17 Oct '20

I have 3 3D numpy arrays of shape 600 x 600 x 200 representing the X, Y, Z directions. The arrays contain dip azimuth, dip magnitude, and edge magnitude calculated from a geophysical dataset. I would like to stack them into a 4D array of shape 600x600x600x3 where azimuth is Hue, dip is Saturation, and edge is Value. *Question*: Would color skimage.color.hsv2rg take this 4D array as input and convert it to RGB? My goal is to then split the 3 R,G,B arrays and load them into a commercial software where I can visualize them using RGB blending? Thanks, Matteo Matteo Niccoli P. Geo., M.sc. Geophysics and data science [image: WordPress] <http://mycarta.wordpress.com/>[image: LinkedIn] <http://www.linkedin.com/in/matteo-niccoli-031aa426>[image: Twitter] <http://twitter.com/@my_carta>[image: Custom link] <https://github.com/mycarta>[image: Custom link] <https://www.researchgate.net/profile/Matteo_Niccoli2>

2 1

how to properly save what I see (on my monitor)
by Scott Bryant 11 Oct '20

11 Oct '20

Hi. I am pretty much a beginner in Python and scikit-image. Thank you for all the great resources! I am using sci-kit image for a hobby project involving images of medieval illuminated manuscripts. Here is my question: How do I save images like I see -- in many of your examples, in matplotlib? For example: 1. The last image of colored, segmented coins in https://scikit-image.org/docs/stable/auto_examples/applications/plot_coins_… 2. Or the three results in https://scikit-image.org/docs/dev/auto_examples/segmentation/plot_marked_wa… 3. Or even the three results in https://scikit-image.org/docs/stable/auto_examples/filters/plot_hysteresis.… . I have some test scripts that successfully save using skimage.io.imsave. But the results like in the examples above don't save the color; and/or the results are very dark (like mostly or all black-looking). I also get a lot of warnings about lossy-ness. I've tried a number of things, including matplot's saving methods. But the results are still what I've described. Any help on this would be super-appreciated! Sincerely, Scott PS I can also save what I see pop up when running a script, hitting the disk symbol and saving. But as you probably know, the results not only have the axes and words and such, but they also are very small and poor resolution (though I'm not confident "resolution" is the right term).

3 2