Find more posts tagged with
Sort by:
1 - 5 of
51
Hi,
I don't know, how images looks like and what you want to detect. For example you want detect water, filed and urban, than I think, it is suitable. For more difficult or special task will maybe be needed new feature extraction operators...
There are few general steps which you must walk through:
1) preprocessing - histogram normalize, resize, denoising etc.
2) create mask - manually select regions which you want to detect
3) generate example set - get features from segment or points + label from mask
4) train model - maybe some optimizations...
5) test model
I don't know, how images looks like and what you want to detect. For example you want detect water, filed and urban, than I think, it is suitable. For more difficult or special task will maybe be needed new feature extraction operators...
There are few general steps which you must walk through:
1) preprocessing - histogram normalize, resize, denoising etc.
2) create mask - manually select regions which you want to detect
3) generate example set - get features from segment or points + label from mask
4) train model - maybe some optimizations...
5) test model
Hello,
I post few examples on http://www.myexperiment.org/users/18932/workflows. I will add some more later. If you have special request, you can write it here...
Best
Vaclav
I post few examples on http://www.myexperiment.org/users/18932/workflows. I will add some more later. If you have special request, you can write it here...
Best
Vaclav
official homepage of extension developers is http://spl.utko.feec.vutbr.cz/en. But work on documentation and tutorials is still in progress. If you have some questions, you can write it here or if you need profi support, contact developers.
Regards
Vaclav