3D+2D+InfraRed Face Recognition
7D: 3D+2D+IR
1. We segment the IR image to identify areas of the UV space of the deformation image that correspond to skin rather than accessories or hair.
 |
|
Original IR Frame
|
 |
|
Hair & Glasses Removed
|
2. Fitting the AFM to the 3D mesh provides a UV parameterization for our mesh that we can use to build a parametric thermal image in the same UV space of the model.
|
|
|
|
|
Infrared Mapped onto 3D Data
|