Compression of simple geometric models using spanning trees

• US 5,825,369 A
• Filed: 07/30/1996
• Issued: 10/20/1998
• Est. Priority Date: 01/16/1996
• Status: Expired due to Term

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Taubin, Gabriel, Rossignac, Jarek Jaroslaw Roman
• Primary Examiner(s): Nguyen, Phu K.
• Application Number:
  US08/688,572
• Time in Patent Office: 812 Days
• Field of Search: 345/440, 345/441, 345/442, 345/443, 345/133, 345/118
• US Class Current: 345/440
• CPC Class Codes: G06T 9/40 Tree coding, e.g. quadtree, octree

Abstract

A computer system stores and transmits compressed simple triangular meshes. The computer uses a data structure that represents a simple triangular mesh in n-dimensional space. The data structure has a table of vertex runs, a table of triangle runs, zero or more marching records, which provide the connectivity information of the triangular mesh. The data structure also has zero or more associated data records that include the geometric information of the triangular mesh. The table of triangle runs and the marching record have information that describes how to construct a triangular mesh (therefore, the polygon vertices and the boundary edges). The table of vertex runs describes a vertex spanning tree that provides additional connectivity information to construct the triangular mesh from the polygon. The associated data record determines the exact position of the triangular mesh in space. The system also includes ways for creating this data structure from a simple triangular mesh, transmitting this information between computers, and compressing and decompressing this transmitted information.

Claims

1. A data structure in a computer memory representing a set of points of a n-dimensional space, the data structure comprising:

a table of runs describing a rooted tree, the rooted tree having one or more nodes, each node being a regular node, a branching node, or a leaf node, each node of the rooted tree represents a point of the n-dimensional space, the table of runs having one or more records, each record representing a run of the rooted tree, a run having a first node that is a leaf or branching node and a last node that is a leaf or branching node and the run connecting one or more regular nodes between the first node and the last node where the connection between every pair of consecutive nodes in the run is an edge, each record having a length of run field in terms of the number of edges in the run, a leaf field indicating if the run ends at a leaf node, and a last run field indicating if the run represented by the record is the last one with the same first node in the rooted tree, the records given by an order of tree traversal with respect to the root node.

2. A data structure, as in claim 1, further comprising an associated data field, where each associated data field is a corrective term, each corrective term being the difference between an actual position of a current point and a predicted position of the current point, where the current point is associated with one of the nodes being called a current node.

3. A data structure, as in claim 2, where the predicted position of the current point is determined by a predictive method that uses one or more ancestor points of the current point, where the ancestor points correspond to nodes that lie on a unique path from the current node to the root node.

4. A data structure, as in claim 1, where the associated data fields are in fixed point format.

5. A data structure, as in claim 1, where the associated data records are compressed.

6. A data structure, as in claim 5, where the associated data records are compressed by any one of the following techniques:

Huffman encoding, Arithmetic coding, Shannon-Fano-Elias coding, or Lempel-Ziv coding.

7. A data structure, as in claim 1, that is used for storing point locations in a computer storage device.

8. A data structure, as in claim 1, that is sent from a computer over a communications link.

9. A data structure, as in claim 1, that is received by a computer over a communications link.

10. A data structure, as in claim 1, where the points represent vertices of an n-dimensional geometric shape, each tree edge corresponding to a respective shape edge of the geometric shape.

11. A data structure for representing a simple triangular mesh in n-dimensional space, comprising:

a table of vertex runs describing a vertex spanning tree, the vertex spanning tree having one or more vertex nodes, each vertex node being a regular vertex node, a branching vertex node, or a leaf vertex node, each vertex node of the vertex spanning tree associated with a vertex of the mesh, the table of vertex runs having one or more vertex run records, each vertex run record representing a vertex run of the vertex spanning tree, the vertex run having a first vertex node that is a leaf or branching vertex node and a last vertex node that is a leaf or branching vertex node and the vertex run connecting one or more regular vertex nodes between the first vertex node and the last vertex node where the connection between every pair of consecutive vertex nodes is a vertex tree edge, each vertex run record having a length of vertex run field given by the number of vertex tree edges in the vertex run, a vertex leaf field indicating if the last vertex node of the vertex run represented by the vertex run record is a leaf vertex node, and a vertex last run field indicating if the vertex run is the last vertex run sharing the same first vertex node, the vertex runs given by an order of traversal of the vertex spanning tree with respect to a vertex root node;

a table of triangle runs describing a triangle spanning tree, the triangle spanning tree having one or more triangle nodes, each triangle node being a regular triangle node, a branching triangle node, or a leaf triangle node, each triangle node of the triangle spanning tree associated with a triangle of the mesh, the table of triangle runs having one or more triangle run records, each triangle run record representing a triangle run of the triangle spanning tree, the triangle run having a first triangle node that is a leaf or branching triangle node and a last triangle node that is a leaf or branching triangle node and the run connecting one or more regular triangle nodes between the first triangle node and the last triangle node where the connection between every pair of consecutive triangle nodes is a triangle tree edge, each triangle run record having a length of triangle run field given by the number of triangle tree edges in the triangle run represented by the triangle run record, and a triangle leaf field indicating if the last triangle node of the triangle run represented by the triangle record is a leaf triangle node, the triangle runs given by an order of traversal of the triangle spanning tree with respect to a triangle root node;

a marching record having a triangle root field and one or more sequences of marching commands, the triangle root field describing how to construct the triangle associated with a triangle root node of the triangle spanning tree, each sequence of marching commands indicating how to construct triangles from one of the triangle runs by advancing to a next vertex along a left run boundary or a right run boundary of the triangle run, the marching commands given by the order of traversal of the triangle spanning tree with respect to the triangle root node; and

an associated data record, the associated data record having one or more associated data fields, each associated data field with information about the position of one of the vertices of the simple triangular mesh, the associated data fields given by an order of traversal of the vertex spanning tree with respect to the vertex root node.

12. A data structure, as in claim 11, where the order of traversal of the vertex spanning tree includes any one of the following:

depth first, breath first, preorder, and postorder.

13. A data structure, as in claim 11, where each associated data field is a corrective term, each corrective term being the difference between an actual position of a current vertex and a predicted position of the current vertex, where the current vertex is associated with one of the nodes being called a current node.

14. A data structure, as in claim 13, where the predicted position of the current vertex is determined by a predictive method that uses one or more ancestors vertices of the current vertex, where the ancestor vertices correspond to nodes that lie on a unique path from the current vertex node to the vertex root node.

15. A data structure, as in claim 11, where the associated data fields are in fixed point format.

16. A data structure, as in claim 11, where the associated data records are compressed.

17. A data structure, as in claim 16, where the associated data records are compressed by any one of the following techniques:

Huffman encoding, Arithmetic coding, Shannon-Fano-Elias coding, or Lempel-Ziv coding.

18. A data structure, as in claim 11, where the order of traversal of the triangle spanning tree is determined by the order of triangle run records in the table of triangle runs.

19. A data structure, as in claim 18, where the order of traversal of the triangle spanning tree includes any one of the following:

depth first, breath first, preorder, and postorder.

20. A data structure, as in claim 11, where the marching records are compressed.

21. A data structure, as in claim 11, where the marching records are compressed by any one of the following techniques:

Huffman encoding, Arithmetic coding, Shannon-Fano-Elias coding, or Lempel-Ziv coding.

22. A data structure, as in claim 11, where each of the marching commands is a bit.

23. A data structure, as in claim 11, further comprising:

one or more additional data records, each additional associated data record having one or more additional data fields, each additional data field having additional information.

24. A data structure, as in claim 23, where the additional information includes one or more of the following:

normals, colors, and textures coordinates.

25. A data structure, as in claim 11, that is used for storing a triangular mesh in a computer storage device.

26. A data structure, as in claim 11, that is sent from a computer over a communications link.

27. A data structure, as in claim 11, that is received by a computer over a communications link.

28. A data structure, as in claim 11, for representing a connected component of a triangular mesh in n-dimensional space having multiple connected components.

29. One or more computer systems for compressing and decompressing a geometric model, each computer system comprising:

a central processing unit for executing processes;

a memory;

a table of vertex runs residing on the memory describing a vertex spanning tree, the vertex spanning tree having one or more vertex nodes, each vertex node being a regular vertex node, a branching vertex node, or a leaf vertex node, each vertex node of the vertex spanning tree associated with a vertex of the mesh, the table of vertex runs having one or more vertex run records, each vertex run record representing a vertex run of the vertex spanning tree, the vertex run having a first vertex node that is a leaf or branching vertex node and a last vertex node that is a leaf or branching vertex node and the vertex run connecting one or more regular vertex nodes between the first vertex node and the last vertex node where the connection between every pair of consecutive vertex nodes is a vertex tree edge, each vertex run record having a length of vertex run field given by the number of vertex tree edges in the vertex run, a vertex leaf field indicating if the last vertex node of the vertex run represented by the vertex run record is a leaf vertex node, and a vertex last run field indicating if the vertex run is the last vertex run sharing the same first vertex node, the vertex runs given by an order of traversal of the vertex spanning tree with respect to a vertex root node;

a table of triangle runs residing on the memory describing a triangle spanning tree, the triangle spanning tree having one or more triangle nodes, each triangle node being a regular triangle node, a branching triangle node, or a leaf triangle node, each triangle node of the triangle spanning tree associated with a triangle of the mesh, the table of triangle runs having one or more triangle run records, each triangle run record representing a triangle run of the triangle spanning tree, the triangle run having a first triangle node that is a leaf or branching triangle node and a last triangle node that is a leaf or branching triangle node and the run connecting one or more regular triangle nodes between the first triangle node and the last triangle node where the connection between every pair of consecutive triangle nodes is a triangle tree edge, each triangle run record having a length of triangle run field given by the number of triangle tree edges in the triangle run represented by the triangle run record, and a triangle leaf field indicating if the last triangle node of the triangle run represented by the triangle record is a leaf triangle node, the triangle runs given by an order of traversal of the triangle spanning tree with respect to a triangle root node;

a marching record residing on the memory having a triangle root field and one or more sequences of marching commands, the triangle root field describing how to construct the triangle associated with a triangle root node of the triangle spanning tree, each sequence of marching commands indicating how to construct triangles from one of the triangle runs by advancing to a next vertex along a left run boundary or a right run boundary of the triangle run, the marching commands given by the order of traversal of the triangle spanning tree with respect to the triangle root node.

30. A system, as in claim 29, where two or more of the computer systems is connected by a communication link.

31. A system, as in claim 30, where the communication link includes any one of the following:

an internet, a intranet, a wide area network, a local area network, a radio frequency link, an infra red link, and a serial communication link.

32. A system, as in claim 30, where the data the table of vertex runs, the table of triangle runs, and the a marching record is transmitted from one or more first computer systems to one or more second computer systems connected by the communication link.

33. A system, as in claim 29, further comprising a compression process for creating the table of vertex runs, the table of triangle runs, and the a marching record from the geometric model.

34. A system, as in claim 29, further comprising a decompression process for creating the geometric model from the table of vertex runs, the table of triangle runs, and the marching record.

35. A compression process comprising the steps of:

building a vertex spanning tree and a triangle spanning tree from a geometric model;

describing the vertex spanning tree with a table of vertex runs;

describing the triangle spanning tree with a table of triangle runs; and

establishing a marching record that defines how to construct one or more of the triangles in the triangle spanning tree while traversing the triangle spanning tree according to the table of vertex runs.

36. A method, as in claim 35, where the vertex spanning tree and the triangle spanning tree are built by the following steps:

assigning costs to one or more edges of the geometric model;

defining the vertex spanning tree as the spanning tree with a minimum total cost; and

building the triangle spanning tree with the edges that are not included in the vertex spanning tree.

37. A method, as in claim 35, where the vertex spanning tree and the triangle spanning tree are built by the following steps:

a. choosing a set of one or more current vertices of the geometric model;

b. selecting all triangles that contain a current vertex and that have not been visited as a current triangulated layer;

c. select as layer edges all edges of the triangles that have not been assigned and that contain at least one of the current vertices, the unselected edges being separating edges;

d. assign each of the layer edges to the triangle spanning tree if the layer edges does not create a loop in the triangle spanning tree, otherwise assign the vertex spanning tree;

e. assign each of the separating edges to the triangle spanning tree if the layer edges does not create a loop in the triangle spanning tree, otherwise assign the vertex spanning tree;

f. select the separating edges as the set of current vertices;

g. repeat steps a through f until there are no more edges; and

h. construct the vertex spanning tree and the triangle spanning tree using the respective assigned edges.

38. A decompression process comprising the steps of:

reconstructing vertex positions;

constructing a bounding loop;

determining relative indices of one or more Y-bounding loop nodes; and

reconstructing and linking one or more triangle strips.