Reflective Imaging Solved by the Radon Transform

This letter is concerned with imaging a 3-D scene from a set of 2-D laser images of backscattered intensity without prior knowledge. The interaction between an electromagnetic wave and a medium can be understood and modeled in different ways. In the configuration considered here, the interactions result in a 3-D projection of the scene. After inspection in the measured signal, the reflection data appear as a sum of point spread functions with or without angular limitations. The approach proposed here consists thus in considering reflection data as an incomplete data set of Radon kind used in conventional computerized tomography. Under this approximation, reconstruction techniques, such as the well-known filtered back projection or the Feldkamp algorithm, can be exploited. The advantage of this approach is to provide isosurfacic reconstructions of the scene from reflective data performed with real-time computation times and without prior information. Simulation results on a real laser tomographic data set attest of the strength and of the relevancy of the proposed method.