Abstract
A new method for reconstructing three dimensional object from an endoscope image is presented. The proposed method uses image shading generated by a light source at the endoscope head. The problem of reconstructing three dimensional shape of an object from its shading has been widely studied in the literature of computer vision research. There have been many significant developments made in this area. However, most of the conventional methods are not applicable to this case, since they are based on the assumption that the light source is far from the object surface. The light source of an endoscope is located at its head, so that it is near the object surface. In this case, the image brightness depends on not only the surface gradients but also the distance from the light source to each point on the surface. To deal with this difficulty, we employ an optical system having a point light source at the projection center, which well approximates the imaging system of the endoscope. In addition, we introduce an equal-distance contour, which is a curve composed of surface points at an equal distance from the light source. Propagating these contours using the image shading, we reconstruct the object surface. We use the level-set method for numerical computation. Experimental results for real medical endoscope images show feasibility of this method.
