Non-linear resistive grid kernel estimator useful in single feature, two-class pattern classification

• US 5,365,472 A
• Filed: 11/26/1993
• Issued: 11/15/1994
• Est. Priority Date: 11/26/1993
• Status: Expired due to Fees

• Current Assignee: the united states of america as represented by the secretary of the navy
• Original Assignee: the united states of america as represented by the secretary of the navy
• Inventors: Szu, Harold, Poston, Wendy L., Rogers, George W., Priebe, Carey E., Solka, Jeffrey L.
• Primary Examiner(s): Mai, Tan V.
• Application Number:
  US08/157,553
• Time in Patent Office: 354 Days
• Field of Search: 364/807, 364/822, 395/10, 395/20, 395/23
• US Class Current: 708/801
• CPC Class Codes: G06F 18/2321 using statistics or function optimisation, e.g. modelling of probability density functions
• CPC Class Codes: G06N 3/0675 using electro-optical, acousto-optical or opto-electronic means

Abstract

A discrete non-linear estimator is provided to establish a probability dety function useful in classification of a feature as belonging to one of two classes. A first and second known feature space associated with respective first and second classes are each converted to a plurality of voltages reflecting corresponding first and second feature space bins. A first and second one-dimensional, non-linear resistive grid has input nodes that continuously receive the voltages reflecting the respective first and second known feature space. Each resistive grid produces an output voltage at each of a corresponding plurality of first and second output nodes. Each of a plurality of voltage comparators are used to compare voltages at corresponding first and second output nodes to output a discrete probability density function that considers corresponding ones of the first and second feature space bins. A display may be used to visually or otherwise provide each probability density function from each voltage comparator.

Claims

1. A device for establishing a probability density function useful in classification of a feature as belonging to one of two classes, comprising:

first means for converting a first known feature space associated with a first class to a plurality of voltages reflecting said first known feature space at a corresponding plurality of first feature space bins;

a first one-dimensional, non-linear resistive grid having a first plurality of input and output nodes, each of said first plurality of input nodes continuously receiving said plurality of voltages reflecting said first known feature space, wherein said first resistive grid produces a first output voltage at each of said first plurality of output nodes;

second means for converting a second known feature space associated with a second class to a plurality of voltages reflecting said second known feature space at a corresponding plurality of second feature space bins;

a second one-dimensional, non-linear resistive grid having a second plurality of input and output nodes, each of said second plurality of input nodes continuously receiving said plurality of voltages reflecting said second known feature space, wherein said second resistive grid produces a second output voltage at each of said second plurality of output nodes;

a plurality of voltage comparators, each of said plurality of voltage comparators for comparing said first and second output voltages from corresponding ones of said first and second plurality of output nodes, wherein each of said plurality of voltage comparators outputs a discrete probability density function that considers corresponding ones of said plurality of said first and second feature space bins; and

means for displaying each said probability density function from each of said plurality of voltage comparators.

2. A device as in claim 1 wherein said known feature space is represented in the optical spectrum and said means for converting is a photoresistor.

3. A device as in claim 1 wherein each of said plurality of voltage comparators is a high-gain operational amplifier.