Mystery Picture

Here is a photograph (untouched, not CGI). When I saw it I went wow! (I knew what it was). I’d be interested to know if anyone (a) KNOWS what it is of (b) can estimate the scale (c) has seen anything like it. If you do know, please post a comment saying so [but please DON’T give the answer]. I plan to release more information daily…
Besides the photo itself there is a serious question. How can you search the web for images like this?
picture0.JPG
and a close-up:
picture1.JPG
[UPDATE – more info: The photograph was taken yesterday by Dr. Judith Murray-Rust.]
[ANSWER: This is, indeed, crystalline water but the scale took us by surprise. The x-axis is ca. 20 cm. This artefact appeared in our bird bath and there appear to be 2 perfect, huge, hexagonal ice crystals (it is possible that they are both sixfold twins, I suppose). The faces are highly planar and specular (we have more pictures).
It is also remarkable that there are two artefacts separated by 10 cm(between centres) which are almost identical. What possible coupling could there be between them – that is the real mysetery.]
Happy Holliday – as I might say to Gemma.

This entry was posted in fun, semanticWeb. Bookmark the permalink.

7 Responses to Mystery Picture

  1. I think I know what it is. But I can just be bluffing…

  2. Does it have anything to do with the crystalline form of a seasonal precipitation, in the seed state?
    db

  3. You can’t search the Web for images like that unless they are properly described, at least given the present state of image-recognition software. This is why librarians should be involved in data curation from the beginning; we realize things like that. Too many domain experts think it’s just a matter of sprinkling magic computer pixie dust.
    (Not you, of course.)

  4. pm286 says:

    (1), (2) ??? I will now accept guesses of scale.
    (3) Yes you can. I had a demo from a Cambridge company which had trained software to look for “red flower”, “two cats” based solely on the pixels. No tags. No metadata. No catalogs. It uses machine learning to recognise features – and they started with domestic ones. I imagine this can spread rapidly to most common objects. Things like the above picture would be more difficult but I am not hopeless.

  5. Random Chemist says:

    Is it Madonna buried under a mound of snow?

  6. So, I was pretty close…
    As to coupling between artefacts, what’s the underlying pattern in the bowl of your birdbath, there could be channels perhaps or a granularity…
    d

  7. pm286 says:

    (6) no obvious coupling
    It seems like a nice case of symmetry -breaking.

Leave a Reply

Your email address will not be published. Required fields are marked *