Results 1 to 7 of 7
  1. #1
    Join Date
    Nov 2011
    Location
    Mudgeeraba, Gold Coast
    Age
    83
    Posts
    178

    Default Is This Worth Considering ?

    With all the threads I see asking to identify different timbers, I just wondered if any thought had been given to an electronic scanning program.

    With 'face recognition' in photography and 'fingerprint recognition' on mobile phones I would think that 'species recognition' from quality photographs would be a possibility.

    Just thought this was worth a line or two, for discussion.

    Cheers, Fred

  2. # ADS
    Google Adsense Advertisement
    Join Date
    Always
    Location
    Advertising world
    Age
    2010
    Posts
    Many





     
  3. #2
    Join Date
    Feb 2006
    Location
    Perth
    Posts
    27,757

    Default

    This gets raised quite a few times and given you have been a member here since 2011 I and a bit surprised you haven't seen it before.

    Even though there have been claims to the contrary the consensus is that 1/2 photos of a tree won't do it.

    If you look at a botanical book on trees you will see that botanists need close up, very clear photos of things like juvenile and mature leaves, bark, bud, cones, flowers and seeds to identify species. Even if they are available (seeds and flowers are not always on show or too high up to get a clear picture) most folks don't know how to indemnify some of these in the first place let alone take a decent photo of them. Even with all this info some trees species cannot be distinguished and require a very finely sanded, polished cross section of grain to confirm identification.

  4. #3
    Join Date
    Oct 2008
    Location
    Bouvard - Western Australia
    Posts
    325

    Default

    I can't see this working, unless it was really a comprehensive programme.
    Two popular timber in WA are She-oak & Jarrah.
    The colour of She-oak varies from off white to orange, depending on soil type, rainfall & area grown. It also goes very dark after a "hot" bush fire has been through the area.
    Jarrah can vary in colour from pink through red, brown to Dark Purple. This again depends on the soil type, rain fall & area growing.

    You are probably more confused than ever now.
    Col
    Chucks are like potato chips....you can't have just one.

    www.bouvardbush.com
    http://www.mandurahwoodturners.com/

  5. #4
    Join Date
    Nov 2011
    Location
    Mudgeeraba, Gold Coast
    Age
    83
    Posts
    178

    Default

    Thanks for your reply's Bob and Col. Thought it might be a longshot, but had to ask the question. Cheers, Fred

  6. #5
    Join Date
    Jun 2004
    Location
    Mareeba Far Nth Qld
    Age
    83
    Posts
    3,069

    Default

    There was a program created by the CSIRO many years back. Jugo Illic was primarily responsible for it's development. Sadly that program is long gone. Even so, it still required an understanding of the properties of wood and the various identifiable characteristics, eg, pore size, arrangement, colour, source, etc. I did locate the elements of that program but were lost with a computer collapse. A major problem is the use of local names which can and do vary widely for the same species. Then there is the problem of what species to include and which species to ignore.

    Jim
    Sometimes in the daily challenges that life gives us, we miss what is really important...

  7. #6
    Join Date
    Nov 2012
    Location
    Brisbane
    Posts
    1,809

    Default

    No, a simple scanning program will not work. As Blackbutt says above, there is too much colour variation and very fine details are required for critical identification of many species.

    Computer identification keys for hardwoods (flowering plants) were around for awhile, in addition to Jugo Ilic's program there was one based on the NSW Forestry Department's data and also one for international woods (Oxford key). They ran on MS-Dos. I had a copy of those programs (think I deleted them only recently) as well as the card identification key. Those keys all required the use of a good 10 to 20 times hand lens and/or a microscope as well as a thorough understanding of wood technology to count the number and patterns of the vessels in the wood as well as the type of other cells present. Last time I tried to use the timber keys I found I had forgotten too much wood tech terminology to be able to (and I have a PhD in botany!)

    Critical id for conifers (pines) is even more difficult as you need a longitudinal section (as in down the tree rather than across it) to identify the patterns in the plates at the end of the tracheids. Definitely microscope work.

    Recognition identification of timbers is of course available and common - as evidenced by the numerous people on the forums who can "identify" from a photograph, or better still from a wood sample, but at this stage requires the capacity of the human brain.

    As BobL says, botanists use many features of a plant to identify it, and sometimes a critical id cannot be provided without a feature/character that is not available.

    In the future, such a program using artificial intelligence should be possible, but the $$ to create it would be hard to find.

    There are also people who still have the knowledge to do timber ids, and they do use the programs, but they charge for the commercial service.

    David

  8. #7
    Join Date
    Mar 2004
    Location
    Brisbane (western suburbs)
    Age
    77
    Posts
    12,095

    Default

    Well, I'm not sure it couldn't be done. We've had numerous programmes for reading histological tissues for yonks, and your average bit of animal tissue is quite a bit more complex than wood micro-anatomy. Face-grain is next to useless, as already mentioned, because there is just way too much overlap between species in colour & grain patterns, but the end-grain patterns are another matter, and they can certainly be used for quite accurate id purposes.

    Just like a skilled operator, the programmes I'm thinking of have to be 'taught' to do their job, and it takes some patience & skill to get them doing their job reliably, but surely, if someone had the time & will (& money) to get it working, you could probably end up with something as reliable as a skilled wood anatomist. It wouldn't come cheap, and you'd need a pretty good scanner to input the data, a bit more computing grunt than your average mobile phone, plus a way of preparing the specimens consistently, so not sure it would be economic or practical for your average weekend warrior......

    Cheers,
    IW

Similar Threads

  1. Worth to buy???
    By Jovian in forum TIMBER
    Replies: 0
    Last Post: 20th November 2012, 01:51 AM
  2. What's a never used MS 880 worth?
    By BobL in forum SMALL TIMBER MILLING
    Replies: 22
    Last Post: 11th March 2009, 10:26 PM
  3. before asking what its worth.
    By weisyboy in forum TIMBER
    Replies: 12
    Last Post: 14th July 2008, 12:17 PM
  4. How much and is it worth it?
    By squashedfrog1 in forum THE GARDEN SHED
    Replies: 4
    Last Post: 22nd January 2007, 09:12 PM
  5. What might this be worth?
    By himzol in forum NOTHING AT ALL TO DO WITH WOODWORK
    Replies: 6
    Last Post: 8th May 2006, 11:12 PM

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •