Back to results
US 4,616,333 AGranted
Data correlation system
AssigneeElscint Ltd.
InventorYair Shimoni
Life of this patent
Estimating the term…
Abstract
A pseudo correlation system and method for determining matches between data defined shapes and templates by obtaining a pseudo correlation value without a successivity of multiplications or determining square roots.
Claims30
A method of determining whether a first shape defined by a first plurality of data matches a predetermined shape, said method comprising the steps of:
- obtaining a selected set of data values (Yi) at a plurality of data points N describing said pre-determined shape,
- obtaining a first set of data values from said first plurality of data using at least the same number of data points as said plurality of data points,
- obtaining a sub-set of data values (Xi) from said first set of data values using said same number N of data points,
- selecting a fiducial value from said selected set of data values and said sub-set of data values, computing a quotient from a dividend and a divisor consisting of summation of differences obtained using said data values from said selected set of data values, said sub-set of data values and said fiducial values, and
- using said quotient to determine whether said first shape matches said pre-determined shape.
2 more elements
A system for determining whether a first shape defined by a first plurality of data matches a pre-determined shape, said system comprising:
- means for obtaining a selected set of data values (Yi) at a plurality of data points N describing said pre-determined shape, means for obtaining a first set of data values from said first plurality of data using at least the same number of data points as said plurality of data points,
- means for obtaining a sub-set of data values (Xi) from said first set of data values using said same number N of data points,
- means for selecting a fiducial value from said selected set of data values and said sub-set of data values,
- means for computing a quotient from a dividend and a divisor consisting of summations of differences obtained using said data values from said selected set of data values, said sub-set of data values and said fiducial values, and
- means for using said quotient to determine whether said first shape matches said predetermined shape.
2 more elements
Dependencies are inferred from the claim text.