Method for matching spatial patterns

• US 6,542,638 B2
• Filed: 02/21/2001
• Issued: 04/01/2003
• Est. Priority Date: 02/21/2001
• Status: Expired due to Fees

• Current Assignee: Shannon Roy Campbell
• Original Assignee: Shannon Roy Campbell
• Inventors: Campbell, Shannon Roy
• Primary Examiner(s): Chang, Jon
• Application Number:
  US09/788,852
• Publication Number:
  US 20020154819A1
• Time in Patent Office: 769 Days
• Field of Search: 382/203, 382/204, 382/195, 382/217, 382/218, 382/209, 382/241, 382/201
• US Class Current: 382/201
• CPC Class Codes: G06T 2207/30176 Document
• CPC Class Codes: G06T 7/0002 Inspection of images, e.g. flaw detection
• CPC Class Codes: G06T 7/33 using feature-based methods
• CPC Class Codes: G06V 10/757 Matching configurations of points or features

Abstract

I present a method for matching the spatial relationships between an input set of feature points and a template set of feature points. A feature point consists of a location in space and a label describing the feature at that location in space. A tessellation over the feature point locations is performed. Next, a search identifies polyhedra that have similar contents, the contents being the angles and labels associated with feature points of the polyhedra. Once a match is found, then appropriate adjacent and neighboring polyhedra are examined. Matching the node labels and angular relationships for a set of appropriate adjacent and neighboring polyhedra extends the volume over which matches exist and significantly increases the certainty that a positive match exists. Because this matching method is based on similarities in node angles and labels, it is independent of transformations in position, rotation, and scale if the angles in the polyhedra created by the tessellation are position, rotation, and scale invariant.

Claims

1. A method for matching spatial patterns between a set of object feature points and a set of template feature points comprising the steps of:

a) performing tessellations on the set of object feature points and the set of template feature points, thereby forming a set of object polyhedra and a set of template polyhedra respectively;

b) comparing, angles and feature points of the polyhedra, between at least one of the object polyhedron, from the set of object polyhedra, and, in turn, each of the template polyhedra, from the set of template polyhedra, to find at least one match therebetween;

c) selecting, an object polyhedron, that lies adjacent to a matched one of the object polyhedra;

d) comparing, angles and feature points of the polyhedra, between the selected one of the adjacent object polyhedra and a similar adjacent polyhedron of a corresponding said matched template polyhedron;

e) repeating steps (c) and (d) until it is possible to determine if a spatial pattern match exists between the set of object feature points and the set of template feature points.

2. The method of claim 1 wherein the object and template polyhedra are formed using a Delaunay triangulation.

3. The method of claim 1 wherein the object and template feature points are obtained from image information including at least one of:

spots, edges, lines, corners, line terminations, line crossings, red points, eyes, wheels, nostrils, and mouths.

4. The method of claim 1 wherein the match criterion is one of:

a selected quantity of the feature points matched, a level of discrepancy between the angles matched, and a quantity of polyhedra matched.