| Boolean operation is a very common approach for constructing complex solid ge-ometries in computer-aided geometry design.Since it was introduced in the 1980s,research works on it are mostly trading off between efficiency and robustness.In earlier times,many Boolean algorithms were developed which are based on BSPTree.However,recently reserchers have lose their enthusiasm on how to apply BSPTree on Booleans on account of its intrinsic flaws.Instead,researchers turned to developing Boolean algo-rithms based on B-rep on which our algorithm is exactly based.Most algorithms require strictly that input meshes have no cavity and boundary edge,which ensures that inputs can serve as boundary representations for some solids.Quite different from the above mentioned,this paper proposes an efficient,robust and adaptive method of Boolean op-eration,which could be applied to non-solid meshes.To make our algorithm is more ef-ficenct,we also developed a fast algorithm on how to make constrained Delaunay trian-gulation.Our algorithm includes some steps as follows.Firstly,we merge inputs meshes into one mesh,then split it into different patches according to non-manifold edges after resolving intersection issue,and identify all cells surrounded by those patches.Next,we propose a method by adding virtual patches to compute the winding number of each cell,which is used to tag the special property of the cell in every input mesh,and con-sequently acquire the correct result of Boolean operation. |
