Measuring frame angles from a digital photo?

plug1n
plug1n Posts: 204
edited January 2009 in Workshop
Anyone know of any software solution for this?

I have no engineering background but I would image cad packagaes must do this sort of thing.

Thx

Comments

  • dennisn
    dennisn Posts: 10,601
    I'd love to help, but am so out of date that I don't even own a digital camera.

    Dennis Noward
  • _Brun_
    _Brun_ Posts: 1,740
    Photoshop has a protractor tool, which will be accurate as long as the photo has as little perspective as possible.

    If you can post the picture or a link to it I could take a look at it this evening.
  • plug1n
    plug1n Posts: 204
    Thanks, I haven't taken any pictures yet

    As a Linux buff, I decided to have a look at gimp which I have installed but never used in anger.

    It has a simple protractor function that measures above and below the horizontal.

    That would work fine as long as the back was exactly horizontal, otherwise you'd have to add or subtract the ground's tilt.

    As a test I took an image from http://www.konaworld.com/09bikes/large/2K9_DRDEW.jpg (which I'm thinking of buying for my son). Then I used the protractor tool to draw a line down the outside of the seat tube and got 73.45 degrees. This would make the frame a 52 or 54 according to their geometry table.

    Of course there is no guarantee that this picture is taken on a true horizontal, but I suppose you could draw a line between the two wheel axles to see.

    Is photoshop any better?
  • julk
    julk Posts: 55
    I did this in photoshop by placing a vertical guide over the photo and rotating the photo until the seat tube etc were parallel to the guide.

    The angle of rotation r gives you the angle of the frame as 90 - r
  • _Brun_
    _Brun_ Posts: 1,740
    Photo is pretty much straight on. Back wheel appears slightly larger than the front, but not by anything significant.

    I got the seat tube angle as 73.6 deg.
  • El Gordo
    El Gordo Posts: 394
    If the photo is anything other than straight on it becomes tricky. If the front is further away than the back or the top further away than the bottom then the apparent angles will get skewed.

    A CAD package and a bit of knowledge about rotating 3D wire-frames will sort it out but I think you'd struggle to be all that accurate. You'd have to use the apparent ovality of the wheels as a benchmark I reckon.

    Why do you want to do it anyway?
  • plug1n
    plug1n Posts: 204
    Why? Well, Cyclefit tells me I need a 72 degree seat-tube angle (but I realise that you can achieve this through putting the saddle ~1cm back on a 73, for example)

    I wanted to measure my 90's Fondriest commuter but also thought it would be good to checkout old frames on Ebay, etc.

Categories