Siddha Ganju, If you would like to work on the OpenCV project I would stick to wrapping OpenCV. I don't think that this is a particularly good project for a beginner as it will invariably involve julia c interop, a c / c++ shim, knowledge of OpenCV data type internals, and julia code generation using python templates. This is not to dissuade you in any way, just a realistic assumption about what you will need to learn to tackle this. If you are looking at this project, I would add that it would be best to look at the matlab bindings as you will have to go through similar hoops to integrate with OpenCV's datatypes given the julia's array layout. Also, the code to generate the matlab bindings is by far the most comprehensible.
Best, Jake On Tuesday, March 11, 2014 3:25:17 PM UTC-4, J Luis wrote: > > Well, given that this thread is under the name of "... OpenCV", why not > using OpenCV itself? The floodfill function does exactly that. > > Terça-feira, 11 de Março de 2014 18:06:46 UTC, Siddha Ganju escreveu: >> >> I have given a thought to creating a Julia wrapper for MagickWand (Issue # >> 69 <https://github.com/timholy/Images.jl/issues/69> ). In this context >> is there any available generator that can help me out? For either >> PythonMagickWand or MagickWand? I found Clang.jl that I can use as a ready >> reckoner for MagickWand. There might be others that I may be unaware of, >> please list them. However I can find none for Python (I might be blind all >> the same). Since Python is object-oriented I consider it to be more viable. >> But that may just be my opinion. >> Then there is another question, is it possible to use PyCall package for >> calling the Python functions (hence PythonMagickWand), I understand that >> some changes will be necessary. Then I can proceed with a pull-request. >> >> Thank you for the help, >> Siddha >> >
