ofxCv - ofxCv::ObjectFinder and scaled images problem


#1

I’m using a downscaled image from my live cam with ofxCv::ObjectFinder (as a face tracker). this works well to get good performance.
However, the Object finder will draw the region of interest in the ‘wrong’ place when drawn on the full sized image. What it the correct way to get round this?

I can’t find any documentation on ofxCv and there is nothing in the examples of use for this.

TIA


#2

“the ‘wrong’ place” it is really generic, it is difficult to help you. Do you mean, that the regions are drawn as they were obtained from the non-scaled image? Do you have some code snippets?


#3

Ahh, just figured it out - I popMatrix, scale and translate, draw contourFinder then pushMatrix.
Seems to work ok for now.

Thanks