Patent US2567008A successfully refreshed from Google Patents.
Patent US2501103A successfully refreshed from Google Patents.
Patent US5561431A successfully refreshed from Google Patents.
US5561431A
Expired / Lapsed
US
⏱️ 0.0 years remaining
Wavelet transform implemented classification of sensor data
Assignee: Lockheed Martin Corp
•
Filing: 1994-10-24
•
Granted: 1996-10-01
•
Expires: 2014-10-24
Abstract
Abstract An improved classifier estimates a classification of a sensed object. Data representing the sensed object are received by a sensor, such as a radar, and transformed into wavelet transform coefficients. A subset of the wavelet transform coefficients are selected, the number of coefficients in the subset being fewer in number than the size of the original data. The subset of wavelet transform coefficients is then used in place of the original data by a classifier that generates the classification of the sensed object. The classifier may be a correlation (profile matching) classifier or a quadratic classifier. In another embodiment of the invention, a wavelet fusion classifier further improves classifier performance by taking data from a single sensor, and transforming it into a reduced subset of wavelet transform coefficients for use with a correlation classifier, and also transforming the data into a reduced subset of wavelet transform coefficients for use with a quadratic classifier. These respective reduced subsets of wavelet transform coefficients are supplied to corresponding correlation and quadratic classifiers. A confidence level is determined for the output of each of the correlation and quadratic classifiers. The output of the classifier having the highest confidence level is then selected as the estimated classification of the sensed object.
Inventors
- 👤 Larry Peele
- 👤 Charles Stirman
Bibliographic Details
Application Number
—
Jurisdiction
US
Data Source
google_patents
Last Synchronized
2026-11-08 08:56
CPC Classifications
G; G01; G01S; G01S7/00; G01S7/02; G01S7/41; G01S7/411; G06; G06F; G06F17/00; G06F17/10; G06F17/14; G06F18/00; G06F18/20; G06F18/285; G06V; G06V10/00; G06V10/20; G06V10/255
Claims Analysis
Total Claims: 24
Claim 1
Independent Claim
376 words
A classifier comprising: input means for receiving sensor data signals representing a sensed object; first transform means for transforming said sensor data signals into a plurality of first wavelet transform coefficients suitable for use in a first classification processing means; first subselection means for generating a first subset of wavelet transform coefficients, said first subset comprising at least one of said plurality of first wavelet transform coefficients; first classification processing means for generating, from said first subset of wavelet transform coefficients, a first classification signal indicative of a first classification of said sensed object, wherein said first classification processing means comprises: means for comparing said first subset of wavelet transform coefficients to each of a plurality of stored known subsets of wavelet transform coefficients, each stored known subset corresponding to one of a plurality of known classifications; and means, coupled to said comparing means, for generating said first classification signal by selecting a most closely matching one of said plurality of known classifications that corresponds to a most closely matching known subset of wavelet transform coefficients; first confidence level means, coupled to said first classification processing means, for generating a first confidence level signal corresponding to said first classification signal; second transform means for transforming said sensor data signals into a plurality of second wavelet transform coefficients suitable for use in a second classification processing means; second subselection means for generating a second subset of wavelet transform coefficients, said second subset comprising at least one of said plurality of second wavelet transform coefficients; second classification processing means for generating, from said second subset of wavelet transform coefficients, a second classification signal indicative of a second classification of said sensed object, wherein said second classification processing means is a quadratic classifier; second confidence level means, coupled to said second classification processing means, for generating a second confidence level signal corresponding to said second classification signal; and means, coupled to said first and second classification processing means and to said first and second confidence level means, for determining which of said first and second confidence level signals indicates a higher confidence level, and for providing as an output of said classifier a selected one of said first and second classification signals corresponding to said determined higher confidence level.
Claim 2
Dependent Claim
376 words
A classifier comprising: input means for receiving sensor data signals representing a sensed object; first transform means for transforming said sensor data signals into a plurality of first wavelet transform coefficients suitable for use in a first classification processing means; first subselection means for generating a first subset of wavelet transform coefficients, said first subset comprising at least one of said plurality of first wavelet transform coefficients; first classification processing means for generating, from said first subset of wavelet transform coefficients, a first classification signal indicative of a first classification of said sensed object, wherein said first classification processing means comprises: means for comparing said first subset of wavelet transform coefficients to each of a plurality of stored known subsets of wavelet transform coefficients, each stored known subset corresponding to one of a plurality of known classifications; and means, coupled to said comparing means, for generating said first classification signal by selecting a most closely matching one of said plurality of known classifications that corresponds to a most closely matching known subset of wavelet transform coefficients; first confidence level means, coupled to said first classification processing means, for generating a first confidence level signal corresponding to said first classification signal; second transform means for transforming said sensor data signals into a plurality of second wavelet transform coefficients suitable for use in a second classification processing means; second subselection means for generating a second subset of wavelet transform coefficients, said second subset comprising at least one of said plurality of second wavelet transform coefficients; second classification processing means for generating, from said second subset of wavelet transform coefficients, a second classification signal indicative of a second classification of said sensed object, wherein said second classification processing means is a quadratic classifier; second confidence level means, coupled to said second classification processing means, for generating a second confidence level signal corresponding to said second classification signal; and means, coupled to said first and second classification processing means and to said first and second confidence level means, for determining which of said first and second confidence level signals indicates a higher confidence level, and for providing as an output of said classifier a selected one of said first and second classification signals corresponding to said determined higher confidence level.
Claim 3
Dependent Claim
14 words
The classifier of claim 1, wherein the sensor data signals are radar data signals.
Claim 4
Dependent Claim
14 words
The classifier of claim 1, wherein the sensor data signals are acoustic data signals.
Claim 5
Dependent Claim
15 words
The classifier of claim 1, wherein the sensor data signals are optical sensing device signals.
Claim 6
Dependent Claim
15 words
The classifier of claim 1, wherein the sensor data signals are infrared sensing device signals.
Claim 7
Dependent Claim
14 words
The classifier of claim 1, wherein the sensor data signals are sonar data signals.
Claim 8
Independent Claim
314 words
A method for classifying an object, comprising the steps of: sensing said object to produce sensor data signals representing said sensed object; transforming said sensor data signals into a plurality of first wavelet transform coefficients suitable for use in a first classification signal generating step; subselecting a first subset of wavelet transform coefficients, said first subset comprising at least one of said plurality of first wavelet transform coefficients; generating, from said first subset of wavelet transform coefficients, a first classification signal indicative of a classification of said sensed object by comparing said first subset of wavelet transform coefficients to each of a plurality of stored known subsets of first wavelet transform coefficients, each stored known subset corresponding to one of a plurality of known classifications; determining which of said stored known subsets of first wavelet transform coefficients most closely matches said first subset of wavelet transform coefficients; and selecting a most closely matching one of said plurality of known classifications that corresponds to said most closely matching known subset of wavelet transform coefficients; generating a first confidence level signal corresponding to said first classification signal; transforming said sensor data signals into a plurality of second wavelet transform coefficients suitable for use in a second classification signal generating step; subselecting a second subset of wavelet transform coefficients, said second subset comprising at least one of said plurality of second wavelet transform coefficients; generating a second classification signal indicative of a second classification of said sensed object by using a quadratic classifier to generate said second classification signal based on said second subset of wavelet transform coefficients; generating a second confidence level signal corresponding to said second classification signal; determining which of said first and second confidence level signals indicates a higher confidence level; and generating a classification signal by selecting one of said first and second classification signals corresponding to said determined higher confidence level.
Claim 9
Independent Claim
314 words
A method for classifying an object, comprising the steps of: sensing said object to produce sensor data signals representing said sensed object; transforming said sensor data signals into a plurality of first wavelet transform coefficients suitable for use in a first classification signal generating step; subselecting a first subset of wavelet transform coefficients, said first subset comprising at least one of said plurality of first wavelet transform coefficients; generating, from said first subset of wavelet transform coefficients, a first classification signal indicative of a classification of said sensed object by comparing said first subset of wavelet transform coefficients to each of a plurality of stored known subsets of first wavelet transform coefficients, each stored known subset corresponding to one of a plurality of known classifications; determining which of said stored known subsets of first wavelet transform coefficients most closely matches said first subset of wavelet transform coefficients; and selecting a most closely matching one of said plurality of known classifications that corresponds to said most closely matching known subset of wavelet transform coefficients; generating a first confidence level signal corresponding to said first classification signal; transforming said sensor data signals into a plurality of second wavelet transform coefficients suitable for use in a second classification signal generating step; subselecting a second subset of wavelet transform coefficients, said second subset comprising at least one of said plurality of second wavelet transform coefficients; generating a second classification signal indicative of a second classification of said sensed object by using a quadratic classifier to generate said second classification signal based on said second subset of wavelet transform coefficients; generating a second confidence level signal corresponding to said second classification signal; determining which of said first and second confidence level signals indicates a higher confidence level; and generating a classification signal by selecting one of said first and second classification signals corresponding to said determined higher confidence level.
Claim 10
Dependent Claim
14 words
The method of claim 7, wherein the sensor data signals are radar data signals.
Claim 11
Dependent Claim
14 words
The method of claim 7, wherein the sensor data signals are acoustic data signals.
Claim 12
Dependent Claim
15 words
The method of claim 7, wherein the sensor data signals are optical sensing device signals.
Claim 13
Dependent Claim
15 words
The method of claim 7, wherein the sensor data signals are infrared sensing device signals.
Claim 14
Dependent Claim
14 words
The method of claim 7, wherein the sensor data signals are sonar data signals.
Claim 15
Dependent Claim
168 words
A classifier comprising: input means for receiving sensor data signals representing a sensed object; transform means for transforming said sensor data signals into a plurality of wavelet transform coefficients suitable for use in a first classification processing means; subselection means for generating a subset of wavelet transform coefficients, said subset comprising at least one of said plurality of wavelet transform coefficients; first classification processing means for generating, from said subset of wavelet transform coefficients, a first classification signal indicative of a first classification of said sensed object, wherein said first classification processing means is of a first type; second classification processing means for generating, from said sensor data signals, a second classification signal indicative of a second classification of said sensed object, wherein said second classification processing means is of a second type that is different from said first type; and selection means, coupled to said first and second classification processing means, for selecting one of said first and second classification signals as an output of said classifier.
Claim 16
Dependent Claim
168 words
A classifier comprising: input means for receiving sensor data signals representing a sensed object; transform means for transforming said sensor data signals into a plurality of wavelet transform coefficients suitable for use in a first classification processing means; subselection means for generating a subset of wavelet transform coefficients, said subset comprising at least one of said plurality of wavelet transform coefficients; first classification processing means for generating, from said subset of wavelet transform coefficients, a first classification signal indicative of a first classification of said sensed object, wherein said first classification processing means is of a first type; second classification processing means for generating, from said sensor data signals, a second classification signal indicative of a second classification of said sensed object, wherein said second classification processing means is of a second type that is different from said first type; and selection means, coupled to said first and second classification processing means, for selecting one of said first and second classification signals as an output of said classifier.
Claim 17
Dependent Claim
120 words
The classifier of claim 13, wherein the selection means comprises: first confidence level means, coupled to said first classification processing means, for generating a first confidence level signal corresponding to said first classification signal; second confidence level means, coupled to said second classification processing means, for generating a second confidence level signal corresponding to said second classification signal; and means, coupled to said first and second classification processing means and to said first and second confidence level means, for determining which of said first and second confidence level signals indicates a higher confidence level, and for providing as an output of said classifier a selected one of said first and second classification signals corresponding to said determined higher confidence level.
Claim 18
Dependent Claim
86 words
The classifier of claim 13, wherein said first classification processing means comprises: means for comparing said subset of wavelet transform coefficients to each of a plurality of stored known subsets of wavelet transform coefficients, each stored known subset corresponding to one of a plurality of known classifications; and means, coupled to said comparing means, for generating said first classification signal by selecting a most closely matching one of said plurality of known classifications that corresponds to a most closely matching known subset of wavelet transform coefficients.
Claim 19
Dependent Claim
15 words
The classifier of claim 13, wherein said first classification processing means is a quadratic classifier.
Claim 20
Independent Claim
143 words
A method of classifying an object, comprising the steps of: sensing said object to produce sensor data signals representing said sensed object; transforming said sensor data signals into a plurality of wavelet transform coefficients suitable for use in a first classification processing means; subselecting a subset of wavelet transform coefficients, said subset comprising at least one of said plurality of wavelet transform coefficients; using a first classification technique to generate, from said subset of wavelet transform coefficients, a first classification signal indicative of a first classification of said sensed object; using a second classification technique to generate, from said sensor data signals, a second classification signal indicative of a second classification of said sensed object, wherein said second classification technique is different from said first classification technique; and selecting one of said first and second classification signals as an output of said classifier.
Claim 21
Independent Claim
143 words
A method of classifying an object, comprising the steps of: sensing said object to produce sensor data signals representing said sensed object; transforming said sensor data signals into a plurality of wavelet transform coefficients suitable for use in a first classification processing means; subselecting a subset of wavelet transform coefficients, said subset comprising at least one of said plurality of wavelet transform coefficients; using a first classification technique to generate, from said subset of wavelet transform coefficients, a first classification signal indicative of a first classification of said sensed object; using a second classification technique to generate, from said sensor data signals, a second classification signal indicative of a second classification of said sensed object, wherein said second classification technique is different from said first classification technique; and selecting one of said first and second classification signals as an output of said classifier.
Claim 22
Dependent Claim
79 words
The method of claim 17, wherein the step of selecting comprises the steps of: generating a first confidence level signal corresponding to said first classification signal; generating a second confidence level signal corresponding to said second classification signal; and determining which of said first and second confidence level signals indicates a higher confidence level, and providing as an output of said classifier a selected one of said first and second classification signals corresponding to said determined higher confidence level.
Claim 23
Dependent Claim
79 words
The method of claim 17, wherein said first classification technique comprises the steps of: comparing said subset of wavelet transform coefficients to each of a plurality of stored known subsets of wavelet transform coefficients, each stored known subset corresponding to one of a plurality of known classifications; and generating said first classification signal by selecting a most closely matching one of said plurality of known classifications that corresponds to a most closely matching known subset of wavelet transform coefficients.
Claim 24
Dependent Claim
25 words
The method of claim 17, wherein said first classification processing technique comprises the step of using a quadratic classifier to generate the first classification signal.
References Cited (Backward Citations)
US5245587A
Cited
US5379268A
Cited
USPTO Annuity & Maintenance Windows
| Annuity Stage | Window Opens | Payment Deadline | Surcharge Period Until | Estimated Fee | Status |
|---|---|---|---|---|---|
| 3.5 Years (Year 4) | 1999-10-01 | 2000-04-01 | 2000-10-01 | $2000.00 | Window Closed / Lapsed |
| 7.5 Years (Year 8) | 2003-10-01 | 2004-04-01 | 2004-10-01 | $3760.00 | Window Closed / Lapsed |
| 11.5 Years (Year 12) | 2007-10-01 | 2008-04-01 | 2008-10-01 | $7700.00 | Window Closed / Lapsed |