From: david@edenroad.demon.co.uk   
      
   On Mon, 21 Feb 2005 20:23:58 +0000, LH wrote:   
      
   >   
   > "David Mitchell"  wrote in message   
   > news:pan.2005.02.21.19.12.18.315901@edenroad.demon.co.uk...   
   >> On Mon, 21 Feb 2005 18:48:05 +0000, LH wrote:   
   >>   
   >>> There are mathematical ways that one could use to correlate the   
   >>> accuracy of the image to the alleged text.  I know enough about image   
   >>> processing to   
   >>> know that if the Air Force wanted to figure out what that document   
   >>> said, they could processes it well enough to get at least pieces of it   
   >>> decoded.   
   >>   
   >> I beg to differ.  Some letters, perhaps.   
   >   
   > Get some letters and it's simply a word puzzle after that.   
      
   So long as you have enough of them.  Otherwise it's just a Rorschach test   
   for believers.   
      
   <...>   
   >> So they probably wouldn't be too keen to do that.   
   >   
   > Exactly.   
      
   So you can't infer anything from their response.  It doesn't mean they're   
   hiding anything.   
      
   > Why not, tell us how you processed the image so anyone can try it and that   
   > would be believable.   
      
   I haven't spent that much time on it, my informal opinion is based on half   
   an hour with The Gimp.   
      
   If you want a formal opinion I could give you something in a couple of   
   weeks; but I don't think you could afford me! ;-)   
      
   > There were no classified image processing algorithms   
   > that I was aware of while I had a clearance, the results of image   
   > classification systems were classified occasionally, but not the actual   
   > algorithms.   
      
   So why don't _you_ spend some time on it.  I'm sure the site author would   
   love you for it.   
      
   >   
   > I've got enough experience and I've seen enough image processing results   
   > to say as an expert that I think they're full of it.  I will agree with   
   > you that you probably would not get complete sentences out of the   
   > document. However, if you could get letters in many of the words, using   
   > the English language and the process of elimination you could decipher at   
   > least sections of it.   
      
   And there are still the other questions to answer: the provenance of the   
   document, and so on.   
      
   --   
   =======================================================================   
   = David    --- If you use Microsoft products, you will, inevitably, get   
   = Mitchell --- viruses, so please don't add me to your address book.   
   =======================================================================   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)