Jump to content


how to implement albedo map?


3 replies to this topic

#1 sky_blue

    New Member

  • Members
  • Pip
  • 5 posts

Posted 02 January 2007 - 04:02 PM

hi,

i want to estimate the albedo map of a face image. i need guidelines on how to do this.
P.S. the image doesn't have to be a human face--it can be anyother object. i just need the general technique.

#2 Reedbeta

    DevMaster Staff

  • Administrators
  • 4782 posts
  • LocationBellevue, WA

Posted 02 January 2007 - 07:42 PM

I'm not sure exactly what you mean by albedo map - I've only heard that term used to mean the same thing as the diffuse color map, which would just be the image itself. Can you clarify what you're looking for?
reedbeta.com - developer blog, OpenGL demos, and other projects

#3 sky_blue

    New Member

  • Members
  • Pip
  • 5 posts

Posted 04 January 2007 - 05:26 AM

The ratio of the amount of light reflected by an object to the amount of incident light is termed as albedo.

I have a list of irradiance value and the corresponding light source direction.
from these information I have to compute surface normal. then I can determine the incident light as "incident light = light source direction dot surface normal".

so if u can tell me the formula for computing surface normal using light source direction, it will be of great help.

#4 Reedbeta

    DevMaster Staff

  • Administrators
  • 4782 posts
  • LocationBellevue, WA

Posted 04 January 2007 - 05:47 AM

So, you know what the albedo of a pixel is, and where the light is coming from, and you want to figure out the surface normal? Well, it depends on the surface's BRDF. If you assume a Lambertian BRDF, for instance, then reflected light is simply light-color * albedo * dot(normal, light-direction). The dot product is of course the cosine of the angle between those vectors, so you could take the arc-cosine. That would establish a cone of possible directions for the surface normal. You can't narrow it down any further than that by looking at one pixel at a time, though you might be able to do better by taking into account neighboring areas of the image, or if you have multiple light sources with linearly independent colors.
reedbeta.com - developer blog, OpenGL demos, and other projects





1 user(s) are reading this topic

0 members, 1 guests, 0 anonymous users