Viewing 14 posts - 1 through 14 (of 14 total)
  • Author
  • #950

    I am interested in using SCiO to recognise various wild fungi, especially to discriminate between edible and poisonous ones.

    I foresee main challenges in sample age and how to tie together scans of the top, stem and underside.

    Luis Cordova

    I had the same idea it would be really good. I have some ideas for an app if you want to discuss more. I think SCIO would have to hook up with fantastic fungi group to get the poisonous mushrooms.

    Francisco del Pino

    previous work that I have access to the creation of a database of drugs (cocaine, cannavis, etc.) and hazardous material.
    It would be interesting to market between police, and concerned parents, etc.
    they see interesting


    Hi Luis and Francisco,
    Great to see common interest in this subject. I will be off to the shops to get some common commercial mushrooms to get a feel for what scan results show.
    I think actual toxins will be too small to identify, but maybe mushroom composition will allow identification.

    Narcotics identification is bound to be a huge deal. The trick will be to have access to known samples without attracting handcuffs!

    Luis Cordova

    Let us know what commons you will be testing, I will be doing the same once I get my tester. I think a way to tell different mushrooms can be done but I need to see the data first. I was thinking we could build an app that says ok to eat (green) and do not eat (red). Maybe even have like info about the mushroom identified like a book.


    Mushrooms are an interesting specimen and has a few cororlations with another project of mine.


    Many kinds of shrooms can be bought at the local market, but its the hallucination  and posison ones that are in short supply as specimens to test.  One model could be to exclude anything that isn’t a verified “safe” mushroom but identifying a morell .vs. a protabella, both safe to eat, could solve the same issue  without the need of the dangerous ones…

    Francisco del Pino

    ello rejsharp

    Until you have on hand SCIO not be able to make the drug database, no problem, you can also make explosives,
    big business to exploit and organize


    Hi Francisco,
    I never thought of explosives! I am sure that would be a great project. Like getting narcotics samples, the problem will be getting reference samples in a safe way. Maybe a cooperation with a testing laboratory or enforcement office?
    Unfortunately I do not have such contacts.

    @ Luis:
    I have scanned three types of mushrooms so far (one tray purchased of each from local shop)
    – King Oyster mushroom Pleurotus eryngii
    – white cultivated mushrooms
    – brown cultivated mushrooms a variety of agaricus bisporus
    As may be expected the King oyster shows quite a difference from the common champignons
    I hope to get some other varieties scanned soon.

    Please remember that CP have warned that the scans from prototype devices may or may not be useful in later collections.
    I am not building huge collections yet, just a few samples to see what may be promising, and to test the software.

    Philippe Laroche

    hello rejsharp,

    when you scanned the mushroom, what part did you scan ?

    do you try the cap, the stem, the underside of the cap, and did one the place seems a better place (more distinct spectrum) ?

    i will try as soon as i have my scio, it would be a good idea to test the inside (after cutting the cap or the stem), because it think a lot of things in a real forest can get glue on the cap.




    Hi Philippe,
    I scanned all three places cap, underside and stem.
    Mostly the scans are very different between the three, even with the same actual mushroom.
    Not enough samples scanned yet, but the underside looks promising. I am worried about cap scans because of variability – we will see.

    It is extremely hot and dry in Austria now, so no wild mushrooms to be found at the moment.

    regards — Roger


    I would expect three different looking scans based on top/bottom/stem scans.  The next step I think CP is going to do for us is give us access to some of the SDK so we can write a program that prompts the user to…  Scan top of mushroom cap…. scan underside of cap… scan stem…  Then using three scans from three different defined areas to do final identification.



    BTW, avoiding jail while researching things like trinitrotoluene (TNT) and ammonium nitrate/fuel oil (ANFO) totally depends on where you live…




    I would guess that if you contact your local authorities, and explain what it is you’re trying to do, that you may be able to gain access to an evidence room, under supervision of course.


    Hi All,

    Mushrooms are my main area of interest and I have Scio and SDK. I have made a couple of efforts towards a model but realize the questions are far too many to approach in a broad way.


    Someone mentioned morel and portabello earlier, but simple broad statements such as “these are both safe”are dangerous. As morel is not safe to eat in a raw state, and must be cooked to remove the hydrazine.


    I have started a collection of wild mushroom scans, as there have been a couple varieties available to do so, but I am not sure what attributes to apply. Can I simply scan as many samples as possible now and apply attributes later? This is the number one question I have.

    In my scans I have taken scans of the three surfaces mentioned above as well as cut surface.


    The mushrooms I have scanned are

    Lactarius deliciosus

    Leccinum scabrum

    Lentinula edodes

    Agaricus bisporus – white and brown


    The season has started so there will be many more





    Would anyone like to co develop a model? I am a mushroom wholesaler, so access to a range of strains, of many ages and origins is in my daily scope.

    Is there a way to share scans? I was hoping for more open source access.

    If not then we can at minimum brainstorm different attributes to apply.


    if interested I’m

Viewing 14 posts - 1 through 14 (of 14 total)
  • You must be logged in to reply to this topic.