Vector correlation system for automatically locating patterns in an image
First Claim
1. For use in an automatic image processing system, a method for locating a position of an object within an image, the image defined by a plurality of image pixels, each image pixel having a known intensity value and a known coordinate point, comprising the steps of:
- determining a template corresponding to an estimation of the object;
translating said template to an abstract template by defining said abstract template using a number of vectors comprising a set of coordinate points and corresponding edge directions, each coordinate point lying on an edge of said template and a corresponding edge direction passing through said coordinate point perpendicular to said edge at said coordinate point, and using vectors along substantially vertical and substantially horizontal extents of said template;
minimizing said number of said vectors;
selecting at least one vector for each different edge of said template only where a sufficiently large angular difference exists between adjacent edges;
converting said abstract template to a ranging template;
displacing said ranging template to each of a plurality of known coordinate point locations within said image; and
correlating said ranging template with said image at each of said locations until a maximum correlation is located, wherein the location corresponding to said maximum correlation indicates the position of the object.
0 Assignments
0 Petitions
Accused Products
Abstract
The present invention relates to a combined approach of image and template correlation, and vector correlation wherein edge detection and pattern searching are joined in a single operation to provide great flexibility without requiring extensive computational facilities. A template is devised from a sparse set of coordinate points that model only the important edges of an object to be located within an image. Different templates are dynamically reconstructed as an estimation of the object location within the image becomes more precise.
48 Citations
3 Claims
-
1. For use in an automatic image processing system, a method for locating a position of an object within an image, the image defined by a plurality of image pixels, each image pixel having a known intensity value and a known coordinate point, comprising the steps of:
-
determining a template corresponding to an estimation of the object;
translating said template to an abstract template by defining said abstract template using a number of vectors comprising a set of coordinate points and corresponding edge directions, each coordinate point lying on an edge of said template and a corresponding edge direction passing through said coordinate point perpendicular to said edge at said coordinate point, and using vectors along substantially vertical and substantially horizontal extents of said template;
minimizing said number of said vectors;
selecting at least one vector for each different edge of said template only where a sufficiently large angular difference exists between adjacent edges;
converting said abstract template to a ranging template;
displacing said ranging template to each of a plurality of known coordinate point locations within said image; and
correlating said ranging template with said image at each of said locations until a maximum correlation is located, wherein the location corresponding to said maximum correlation indicates the position of the object. - View Dependent Claims (2, 3)
determining a displacement defined from said coordinate points of said vectors and extending along corresponding edge directions;
positioning, for each vector of said abstract template, a pair of coordinate points spaced from said coordinate point for said vector along said corresponding edge direction, a positive coordinate point corresponding to a positive value of said displacement and a negative coordinate point corresponding to a negative value of said displacement.
-
Specification