System and method for distinguishing objects

• US 5,754,694 A
• Filed: 06/07/1995
• Issued: 05/19/1998
• Est. Priority Date: 06/30/1992
• Status: Expired due to Term

• Current Assignee: Texas Instruments, Inc.
• Original Assignee: Texas Instruments, Inc.
• Inventors: Villalba, Michael Joseph
• Primary Examiner(s): COUSO, YON JUNG
• Application Number:
  US08/478,710
• Time in Patent Office: 1,077 Days
• Field of Search: 382/224, 382/226, 382/227, 382/228, 382/209, 382/103, 348/135, 348/113, 348/117
• US Class Current: 382/226
• CPC Class Codes: G06F 18/24323 Tree-organised classifiers

Abstract

A method and system are provided for distinguishing objects. Signals representative of the objects are input (63). From the signals, measurements of multiple types are acquired (64). At least one of the measurement types is continuously variable. In response to the measurements, respective amounts of mutual information of the measurement types are determined (70). In response to the respective amounts of mutual information, at least one measurement type is selected for evaluation in distinguishing the objects (72).

Claims

1. A system for distinguishing objects, comprising:

inputting circuitry for inputting first signals representative of the objects;

acquiring circuitry coupled to said inputting circuitry for acquiring measurements of a plurality of measurement types from said first signals, at least one of said measurement types being continuously variable;

determining circuitry coupled to said acquiring circuitry for determining respective amounts of mutual information of said measurement types in response to said measurements;

selecting circuitry coupled to said determining circuitry for selecting, in response to said respective amounts, a measurement type of said measurement types, and;

ordering circuitry coupled to said selecting circuitry for ordering each of said selected measurement types in an n-ary decision tree having branches for evaluation in distinguishing the object, where n is at least 3 and wherein said branches are only directly connected to a parent of said branches, adjacent branches being separated by a threshold value of said measurements.

2. The system as in claim 1, wherein said system further comprises evaluating circuitry coupled to said selecting circuitry for evaluating, in response to said selected measurement type, at least one measurement from a set of second signals from a location for distinguishing at least one of the objects as being represented in said set.

3. The system as in claim 2, wherein said evaluating circuitry is operable to distinguish said at least one of the objects with a specified reliability.

4. The system of claim 1, wherein said ordering circuitry is further operable to quantized each of said selected measurement type to form quantized measurement types being continuously variable, such that said decision tree includes each of said quantized measurement types for evaluation in distinguishing the objects.

5. The system of claim 1, wherein said system further comprises circuitry coupled to said ordering circuitry for traversing a path of said decision tree, such that multiple measurements from a set of second signals from a location are evaluated for distinguishing at least one of the objects as being represented in said set.

6. The system as in claim 1, wherein said system further comprises establishing circuitry coupled to said inputting circuitry for establishing, in response to a representation of the objects in said first signals, initial probabilities of the objects being represented in a set of second signals.

7. The system as in claim 6, wherein said establishing circuitry is operable to establish said initial probabilities in response to a proportionate representation of the objects in said second signals.

8. The system as in claim 1, wherein said measurement types are variably characteristic of the objects.

9. The system as in claim 1, wherein at least one of said measurement types is discretely variable.