Re: texture synthesis code suggestions



Anchor wrote:
I am looking for some dust-busting texture synthesis code to fill in
missing image data caused by dirt or dust on film.

Suggestions?

Texture-synthesis by non-parametric sampling. Cool results if you want to spend the CPU time on it. I'm playing around with a Photoshop filter implementation of the algorithm... if you like you could send me your picture and I will be able to tell you if this algorithm is suitable or not.


Cheers,
Nicholas Sherlock
.



Relevant Pages

  • PSN to Permutation Help Wanted
    ... If you read every single byte sent to usenet, you'll know that I've been playing around with 6/49 lottery forecasting using systems based on magic squares. ... Extrapolating from the data I have, I think it might be possible to build a system which produces a 5-match approximately every 1000 draws. ... Part of the process of building such a system involves some extreme tuning for wrongliness, and to that end I need a algorithm for generating every permutation of a large pool of numbers indexed back to its permutation serial number, in the same way as Duncan published an algorithm for generating combinations from combination serial numbers. ...
    (rec.gambling.lottery)
  • An Extension to the Travelling Tournament problem
    ... annealing algorithm. ... In this situation I have to ensure that when one team is playing at ... and that one plays in division 3 and the other in division 2. ... My algorithm starts with a feasible schedule, ...
    (comp.ai)
  • E[ln(d(x))]
    ... My first post here: hope it's an appropriate place to ask this question. ... I'm playing with a likelihood function for use in an EM algorithm. ...
    (sci.stat.math)
  • Re: the Mathematics of Chess
    ... I.N. Galidakis wrote: ... Chess has been demonstrably modeled completely, ... champion and at any playing moment there is a perfectly well-defined evaluation ... your algorithm to produce the next note with a random midi value from 0 to 127. ...
    (sci.math)
  • Branch and bound ideas needed
    ... I have been playing with a branch and bound algorithm from a paper ... and I am looking for some ideas and pointers to some background info. ... I have tighter bounds but I can't even handle this simple case so lets stick ...
    (sci.math)

Quantcast