Data processing: measuring – calibrating – or testing – Measurement system in a specific environment – Mechanical measurement system
Reexamination Certificate
2003-01-07
2004-04-27
Shah, Kamini (Department: 2863)
Data processing: measuring, calibrating, or testing
Measurement system in a specific environment
Mechanical measurement system
C356S005090, C356S028500
Reexamination Certificate
active
06728645
ABSTRACT:
BACKGROUND OF THE INVENTION
The present invention relates to a method and system for the remote assessment of the vibrating state of an object in general, and to a method and system for target detection, recognition and identification using a Doppler Laser Radar (LADAR) device in particular.
Target acquisition has become an important task in the modem battlefield and future battlefield will demand an increased ability to acquire, transmit, process, disseminate and utilize surveillance and target acquisition information. Yet, because modem weapon systems can operate these days at long distances using sophisticated camouflage and concealment tactics, positive target identification has become actually more and more difficult to implement.
One relatively new source of intelligence information is remote monitoring of the battlefield, using Remotely Monitored Sensors (REMS).
REMS work on the principles of detection of an outside stimulus, logic processing of that stimulus and transmission of a coded signal to a readout device. The tactical unattended ground sensors used by REMS may include sensors, which operate on magnetic, seismic, acoustic, electromagnetic and audio detection principles.
Other attempts to acquire and characterize targets from long distance include electro-optical methods, both passive and active. Among the active optical techniques, it has been known that Doppler LADAR either continuous-wave (CW) or pulsed, can be used to measure distance to objects, the velocities of moving objects and to detect characteristic pattern of the vibration frequencies of the body of an active vibrating target.
Electro-optical systems for detecting the vibration frequencies of an object are referred as Doppler LADAR systems. Updated prior art with regard to Doppler LADAR systems and their method of operation can be found in U.S. Pat. Nos. 5,847,816 and 5,847,817 both to Zediker, et al., U.S. Pat. No. 5,867,257 to Rice, et al. and in the relevant literature such as e.g.: in
Procceding of SPIE
4035, 436 (2000); P. Gatt, et al. “Micro-Doppler Lidar Signals and Noise Mechanisms; Theory and Experiment” in
Procceding of SPIE
4035, 426 (2000); V. N. Glazov, et al. “Analysis of Doppler Laser Radar for identification of dynamic object” in
Procceding of SPIE
4035, 311 (2000); J. F. Fontannella, et al. “Wavelength selection for long range laser vibrating sensing” in
Procceding of SPIE
3380, 107 (1998); B. Lyons, et al. “Upgrades Defense Laser Target Signature Code for the Evaluation of Advanced LADAR Technologies” in
Procceding of SPIE
3380, 164 (1998) and S. H. Hannon, et al. “Agile Multiple Pulse Coherent Lidar for Range and Micro-Doppler Measurements” in
Procceding of SPIE
3380, 259 (1998).
In a typical Doppler LADAR system optical detection starts with illuminating a vibrating or moving target with an intermediate frequency (IF) modulated coherent laser beam of a LADAR. The surface vibrations of that active target further modulate the impinging IF modulated LADAR laser beam frequency, a part of which is reflected back towards the LADAR detector where it interferes with a local reference beam.
The resultant interference products contain a vibration frequency modulated IF component, where the modulation is by the exact Doppler frequency shift of the reflected beam with respect to the reference beam. This modulated IF component is later filtered out in the receiver, to be further analyzed by the LADAR signal processor to extract the vibration frequency features.
Because of improvements in hardware of LADAR systems, e.g. as the transition from the gas laser to the solid state laser, or the use of light pulslets, Doppler LADARs are capable of measuring a surface displacement velocity within a resolution as small as 10 micrometers per second and can thus be used to classify and identify targets based on the vibration signature of a target at a high degree of fidelity at ranges of over 20 km (W. Otaguro, et al.).
Yet the method of processing of the vibration spectrum of a target has little changed over the years and is mainly based on the extracting of the frequency power spectrum of the modulated IF component using the Fourier transform (FT).
The main drawback of fixed windowed FT is that the spatial and frequencies resolutions are fixed and a local feature having a frequency band cannot be located with a precision higher than the frequency width of the window function.
Besides, FT is appropriate for analyzing stationary or quasi stationary signals, but when a signal becomes non stationary or transient because e.g. drift, trends, abrupt changes or beginning and ends of events, information in regard to the order of the events is lost, i.e. it is not known when a particular event took place.
In order to overcome this difficulty of the FT analysis, the technique of short time Fourier transform (STFT) was introduced by Denis Gabor (1946) in which small section of a signal are analyzed each at a time by moving along different sections the signal windows which have different sizes.
Even so, because spatially unresolved vibration signatures, e.g. of a vehicle, depend on range and aspect angle, on engine condition, on the velocity of the vehicle and on the exact location of the laser beam on target, identification becomes complex and requires a more flexible approach of data analysis, so it will be possible vary the window “shape” in addition to the change of its size (as in the STFT), in order to make the most favorable trade-offs between time and frequency resolutions of the signal.
The technique of wavelet packet analysis of signals provides such approach and its application in Doppler LADAR vibrometry is described e.g. in G. A. Harrison, et al. “
Application of wavelet and Wigner analysis to gas turbine vibration signal processing
”, SPIE 3391, 490, (1999).
When it comes to an automatic identification and classification of targets in accordance to their vibration spectrum, two more difficulties are encountered:
The first one is the assignment of an immutable vibration signature to an object regardless its momentarily vibrating state. This is demonstrated in
FIG. 1
to which reference is now made.
Shown in
FIG. 1
are the FT frequency power spectra of the vibration of a door of a truck at three different RPM values of the car's engine, which were measured by a commercial Doppler LADAR vibrometer.
As is seen, although the FT frequency power spectrum of the truck differs in the three RPM values, the spectra include some characteristic features which arc invariant and only these features should be utilized for the positive identification of the car type regardless its vibrating state.
In the example shown in
FIG. 1
these characteristic features include e.g. the power spectrum band structure in the frequency intervals marked as a, b and c.
The second difficulty is demonstrated in
FIG. 2
to which reference is now made.
Shown in
FIG. 2
are the frequency power spectra of car A and car B, which should be distinguished in accordance to their vibration spectra.
As can be seen in
FIG. 2
, there are at least three frequency bands (marked as a′, b′ and c′) in which the differences between the spectra of the cars are mostly pronounced, while the other part of the spectra are relatively irrelevant to the differentiation between the cars.
Besides, it is apparent that the spectral resolution needed to differentiate among the cars in accordance to frequency band a′ is greater than the resolution which is needed to resolve the differences between the cars in accordance to frequency band c′.
I.e. it will be wasteful to analyze the vibrations in frequency band a′ and in frequency band c′ with the same resolution (the same degree of Fourier approximation of the signal).
Both difficulties reflect nowadays limitations of an automatic method for recognition and classification of a suspected target in Doppler vibrometry, wherein it is needed to compare an immutable feature of a vibration of a target with thousands of vibration spectra of objects stored in a da
Kaftory Ran
Kozlov Inna
Peled Shimon
Zeiler Eitan
Zlotnick Zvi
Electro-Optics Research & Development Ltd.
Friedman Mark M.
Shah Kamini
LandOfFree
Method and system for automatic identification of objects... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method and system for automatic identification of objects..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method and system for automatic identification of objects... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3230420