Efficient slicing procedure based on adaptive layer depth normal image

In layered modeling for rapid prototyping of products, compromising slicing accuracy and time is a critical issue. Based on adaptive Layer Depth Normal Image (LDNI), this paper proposes an efficient algorithm to achieve this compromise for complex Constructive Solid Geometry (CSG) models. First, each primitive at the tree leaf is converted into an adaptive LDNI solid whose Boolean operation can be performed efficiently. Then, a layered model is constructed directly from the Booleaned LDNI solid since it is actually a set of a layered and ordered point cloud. In addition to speed, efficient use of memory is also taken into account in design of the LDNI-based slicing algorithm. The capability and efficiency of this slicing algorithm are demonstrated by examples.

► An efficient slicing algorithm for CSG solids is proposed based on LDNI.
► A solid is converted into an LDNI representation which is a layered and ordered point cloud.
► Boolean operation is performed on 1D segments in LDNI.
► Loops are constructed efficiently according to three simple rules.
► Mesh with 2 million triangles can be sliced in 2.5 s.