This paper proposes an overlap removal algorithm for an areal 3D scanner system. An areal 3D scanner generates a range image with a single scanning process. Multiple range images for a subject have overlap regions. The proposed algorithm uses a grid representation to store and process the range images unlike other approaches. The grid representation inherits the topological relations of the image points that are corresponding to imaging elements of the camera. The proposed algorithm consists of four major steps. At the first, it deletes points worse than their alternatives. The next step is to revive the deleted points in lake and gulf shape regions to make smooth boundaries of the range image. The third step is to delete points in island or cape shape regions. The last step is to delete points in other range images of which alternatives are in the considering range image.