Importance

Loading Importance Indicators...

Abstract

A uniform distribution of affine invariants (3-tuples) is produced for a plurality of one or more three-dimensional objects. Each of the three-dimensional objects, capable of a plurality of affine transformations, is defined by a set of object points (feature points) selected from an object feature domain. By selecting one or more five-point tuples of the object points, four of the object points divide the object feature domain into a region arrangement of fifteen regions while the fifth point of the five-point tuple lies in one of the fifteen regions and each of the five-point tuples further defines each of the fifteen regions as one of five reentrant polyhedral (RP) arrangements or one of ten non-reentrant polyhedral (NRP) arrangements. A five-point tuple is said to belong to class i if the fifth point of the tuple resides in the i-th of the 15 regions defined by the first four points. A tagger identifies each of the five-point tuples as having one of the arrangements with one of the regions containing the fifth point of the five-point-tuple. During a knowledge accumulation mode, and using either synthetically generated (Monte-Carlo simulation) or real data, an equalizer accumulates knowledge about occupancy patterns incurred by five-point tuples belonging to each of the 15 classes and then derives the necessary remappings that will result in an expected uniform distribution over the range of invariants for all produced invariants. After the knowledge accumulation phase, the equalizer enters its redistribution mode the end-product of which is the creation of a new distribution of affine invariants (3-tuples) that is uniform over the range of invariants. A stacker then stacks the remapped regions to create a complete uniform distribution for all the affine invariants that the transformer produces. The system produces a table which associates combinations of object points (five-point tuples) to affine invariants (3-tuples); the table can be used as the core of an indexing-based store and retrieve system, it can be used to determine and identify over-represented constructs of object points, to determine potential interaction for ligand-receptor complexes, etc.

Loading the Abstract Image...

First Claim