Request to remove Lena as a sample image

Hi all,

Is anyone on here capable of removing Lena as one of the default sample images bundled in with ImageJ/Fiji? Nature Publishing Group no longer accept that image as an appropriate test image in their publications, and Matlab removed it from their default images a few versions back.

Thanks, and happy to discuss further if necessary :slight_smile:

Siân

8 Likes

I removed the Lena sample image from the latest ImageJ daily build (1.52o17). Macros and scripts that use this image will continue to work,

10 Likes

Thank you Wayne, you’re a hero!

2 Likes

may I ask why?
thanks

Of course! Here’s a note on the matter that they published at the end of last year:
https://www.nature.com/articles/s41565-018-0337-2

1 Like

thank you.
Very strange argument imho, but ok.

I guess from my perspective (as a woman in programming, image analysis, physics etc) things like the Lena image add to my constant feeling that computer science is a boys’ club with in-jokes about Playboy models where women aren’t taken seriously. It’s a small thing to change, but hopefully if enough small things change then women will feel a bit more welcome in the community :smile:

10 Likes

Ok, but for me it was just the opposite use a woman instead of a man picture in data/image science as state-of-the-art image for comparison and algorithm.
Take women so much seriously to use a picture of a woman to compare algorithms.

I know people could talk and argue years about that :slight_smile:

Dear Siân,
isn’t it “the greater the odds, the greater the glory” ?
(translated from: " À vaincre sans péril, on triomphe sans gloire" - Le Cid - Corneille).
I guess being a man I can’t allow myself to comment directly your text, and this even if I have only daughters as kids (which I try to convince as much as I can to take as well a path in science).
Nevertheless, having a degenerative disease, I have been the collateral victim of some unfair decisions of local politicians. It’s vice-president (29 years of teaching experience in Mathematics) tried to demonstrate me mathematically how wrong I was using their measured data:
http://bus263.free.fr/download/2015-01-12_elements_statistiques_complementaires.pdf
So I simply regenerated the numerical data from their graphs (using ImageJ of course!) and wrote him an extremely detailed demonstration using the following obtained graphs:
http://bus263.free.fr/bus263_12.html
and showing him point by point how on the contrary he/they was/were wrong…
And today, it is more than a year now that I’m (still) waiting for his corrections…(It seems Phoenix isn’t his nickname.) :smile:
Also, if the (from my point of view at least) very soft picture of Lena is shocking some people, it is of course better to get rid of it.
But in the same time, such censorship makes me remember a WWE match I saw by a friend (during my post-doctoral experience in the USA) were some ultra-violent scenes were fully shown. But in the same time, they took really care of very carefully hiding the tattoo of one of the boxer representing a naked woman.
I may very probably be kind of weird, but as far as I’m concerned, I was way more shocked by the shown violence scenes rather than a very small tattoo that I would probably even not have noticed if it wouldn’t have been blurred.
So all is relative!
I wish you a very nice week-end.
My best regards,
Philippe

Context is important - I have spent a large part of my career in veterinary schools, which for 30 years at least have been increasingly “girls’ clubs” (something like 75-90% of undergrads are female, around the world: we are still trying to work out why). So it’s always a huge contrast when I go to engineering-type groups where the split is reversed. I can see how such an image in the comp sci / engineering context is problematic, given its origin and history.

I show Lena, once, to my undergrads. I give them an explanation of the background and now, thanks to this discussion, a link to the NPG comment. There’s a huge canon of Lena-copies in the literature: I want my students to know what they are looking at and not be surprised, ignorant or naïve, when they come across a copy, and to understand why people used to use Lena, and now they don’t.

4 Likes

Would it be an alternative to add e.g. the Mandrill image (as stated in the NPG article). Due to its appearance it might be a good test image, specifically for feature preservation in filtering. Obviously, everybody can get a version or use alternatives, but since it is suggested as an alternative already (and I think it is a better one) it would be good to have a sample link, so that everyone who wants to use it in ImageJ uses the same (size, resolution, color panel,…). Here’s the link to the USC image database for the Mandrill image: http://sipi.usc.edu/database/database.php?volume=misc&image=10#top