Abstract

We consider the problem of estimating the 3D shape and re?ectance properties of an ob ject made of a single material from a calibrated set of multiple views. To model refiectance, we propose a View Independent Refiectance Map (VIRM) and derive it from Torrance- Sparrow BRDF model. Re?ectance estimation then amounts to estimat- ing VIRM parameters. We represent ob ject shape using surface trian- gulation. We pose the estimation problem as one of minimizing cost of matching input images, and the images synthesized using shape and re- fiectance estimates. We show that by enforcing a constant value of VIRM as a global constraint, we can minimize the matching cost function by iterating between VIRM and shape estimation. Experiment results on both synthetic and real ob jects show that our algorithm is efiective in re- covering the 3D shape as well as non-lambertian refiectance information. Our algorithm does not require that light sources be known or calibrated using special ob jects, thus making it more fiexible than other photomet- ric stereo or shape from shading methods. The estimated VIRM can be used to synthesize views of other ob jects.