Abstract

Conventional photometric stereo has a fundamental limita- tion that the scale of recovered geometry is limited to the resolution of the input images. However, surfaces that contain sub-pixel geomet- ric structures are not well modelled by a single normal direction per pixel. In this work, we propose a technique for resolution-enhanced pho- tometric stereo, in which surface geometry is computed at a resolution higher than that of the input images. To achieve this goal, our method first utilizes a generalized reflectance model to recover the distribution of surface normals inside each pixel. This normal distribution is then used to infer sub-pixel structures on a surface of uniform material by spatially arranging the normals among pixels at a higher resolution according to a minimum description length criterion on 3D textons over the surface. With the presented method, high resolution geometry that is lost in con- ventional photometric stereo can be recovered from low resolution input images.