Color-Based Edge Detection on Mesh Surface

1 ABSTRACT Edge detection is one of the fundamental techniques and can be applied in many places. We propose an algorithm for detecting edges based on the color of a mesh surface. To the best of our knowledge, we are the first to detect edges on mesh surface based on the color of the mesh surface. The basic idea of our method is to compute color gradient magnitudes of a mesh surface. To do that, the mesh is split along the intersection of surfaces into some segments. Then, segments are voxelized and assigned a representative color by averaging colors at boundaries between voxels and mesh faces. Artificial neighbors are created for completeness and 3D canny edge detection is applied to the resulting 3D representation. Lastly, additional intersections are added by looking at the intersection of two surfaces. Figure 1 shows the results of our method.