Ask Your Question

Does createLBPHFaceRecognizer() uses spatially enhanced histogram?

asked 2016-02-10 13:32:10 -0500

Raafat Salih gravatar image

updated 2016-02-10 13:34:31 -0500

Hello, I am asking about the createLBPHFaceRecognizer() function, how does it really works? in its parameter there are a number representing sampling point and a radius. As far as I know, the input face image should be divided into regions . Then a histogram should be calculated from each region, and then all these histograms are combined forming one global histogram representing the input face image. This global histogram called spatially enhanced histogram. My question is that did this function do all of that? or it just treat the face image as one region and find one histogram representing the image. And if it does, then how many regions are there? and can i change the number of regions? Thanks

edit retag flag offensive close merge delete

1 answer

Sort by ยป oldest newest most voted

answered 2016-02-10 16:50:39 -0500

berak gravatar image
  • did this function do all of that? - yes.

  • And if it does, then how many regions are there? - have a look here . the image is divided into grid_x * grid_y patches (8x8 by default), and sure you can try to change those numbers. if you increase the grid count, your histograms will get more 'sparse' (more patches == less pixels per patch) and more 'local', and overall the descriptor(feature) will get longer. just try ;)

edit flag offensive delete link more



Oh, thank you very much. I didn't see the grid_x and y parameters Thanks again fro your quick response.

Raafat Salih gravatar imageRaafat Salih ( 2016-02-11 00:26:30 -0500 )edit

Question Tools

1 follower


Asked: 2016-02-10 13:32:10 -0500

Seen: 285 times

Last updated: Feb 10 '16