3D reconstruction of specular surface via a novel structured light approach

Abstract

3D measurement of specular surface is a difficult issue for optical-based structured light techniques. Due to strong reflectivity of target surface, the projected patterns are usually invisible to the camera. In this work, we present a novel structured light sensing approach by exploiting the high dynamic range imaging techniques. The target surface is scanned by the structured light device multiple times with various exposure levels. And then, images of the same pattern index are fused together to get a high dynamic range radiance map. By compressing the radiance map to the image space, a new set of images with more homogenous image intensity distribution and well exposure quality can be obtained. And finally, the 3D reconstruction is applied to these reconstructed images. A stainless steel sheet with very strong reflectance is experimented. And the results show that, the specular surface can be well reconstructed with satisfied accuracy.