[GRASS-dev] Report #2, GSoC Image Segmentation

Hi all,

Completed:

I spent the week learning and implementing the “segmentation library” which deals with tiling large rasters to balance memory constraints with disk I/O. It is totally different from image segmentation, the task of the program, so hopefully the dual meaning of the word won’t cause too many headaches! There is a lot of good information in the GRASS programmer’s manual, that has been good reading.

The data I/O portion is coded, using both imagery groups and segmentation data storage. It compiled today, but quits half way through. Hopefully I’m just missing some small syntax error.

Tasks for next week:
Get I/O to work.
Write the basic image segmentation algorithm.

Blocking:
Nothing yet, there are some bug(s?) preventing the program from running. But I’ve just now figured out where the problem is and need some time to continue looking for the solution.

-Eric