Image analysis – Pattern recognition – Feature extraction
Reexamination Certificate
2000-09-28
2004-05-04
Mariam, Daniel (Department: 2621)
Image analysis
Pattern recognition
Feature extraction
C382S128000, C382S159000
Reexamination Certificate
active
06731804
ABSTRACT:
FIELD OF THE INVENTION
The invention concerns a remote sensing system and method for identification of chemical and biological materials dissolved in liquids that uses microwave radiation to stimulate thermal luminescence of contaminants to develop a fingerprint absorption profile of these contaminants.
BACKGROUND OF THE INVENTION
This invention relates to a system for remotely sensing and identifying contaminants in liquids, by using microwave irradiation to stimulate thermal luminescence of contaminants and a combination of spectroscopy and neural network technologies to classify their fingerprint spectra.
Liquid quality analysis systems and methods for identification of contaminants are well known. For example, detection and classification of chemical impurities in a water source are required by most municipal, industrial, and military operations. Accurate real-time detection and classification of hazardous pollutants in waste water discharges and water supplies are necessary for timely mitigation of hazards by any combination of process modification, purification, or rejection of its use or consumption, e.g., when the presence of a particular contaminant is known, specific remedial measures may then be implemented.
One technique for measuring volatile substances in liquids is the “purge and trap” method. In this technique organics are captured by passing the output of a sparging (bubbling) of an inert gas through a sample volume of the liquid being tested through a trap. This can be a lengthy process, especially if a gas chromatograph is used to measure the various concentrations of contaminants. Gas chromatography is also a maintenance-intensive technique. Doyle, et al, U.S. Pat. No. 5,421,194, describe a sparging-infrared (IR) liquid analyzing system which overcomes these drawbacks but still requires that a mechanism be provided for an inert gas to be bubbled through a sample volume of the liquid being analyzed and a condenser be provided between the sparging vessel and the IR gas cell to reduce water vapor interference with measurements. Their approach is complex.
Another sparging technique, developed by the Du Pont Corporation, sprays a water sample through a volume of inert gas. This approach requires collecting and recirculating the sprayed sample through the spraying process step until all the gas is saturated and consequently implementations of the technique perform in less than real-time. Further, this technique does not address the overlap of water vapor absorption bands with contaminant absorption bands and therefore cannot be used to detect many solutes, although it met Du Pont's needs.
Such liquid quality analysis systems and devices are not completely satisfactory and, in practice, are complex to operate and maintain, do not provide remote, real-time contamination identification, and can be inapplicable to a large number of possible solutes.
The present Applicants have devised, embodied, and tested this invention to overcome these shortcomings of the state of the art and to obtain further advantages.
SUMMARY OF THE INVENTION
The present invention achieves a system and method that provide ongoing remote sampling and analysis of liquid quality in real-time, directly from the liquid source and that accurately identifies chemical and biological materials (CBMs) dissolved in the liquid.
In addition, the present invention achieves a system and method that are accurate and reliable for specific expected contaminants.
The present invention provides a real-time pseudo-active remote sensing system and associated method that learn the features of expected contaminants of liquid sources to be monitored by using a neural network that incorporates the features of these specific anticipated contaminants in advance of monitoring activities. The neural network is then used to match the features of detected contaminants with the features already learned.
In a particular embodiment, called a Thermal Luminescence Water Monitor or TLWM, a water-solute sample is drawn and held in a cell structure designed with an infrared transmission window. The cell structure is coupled to a magnetron or klystron source for generating an energetic microwave beam. Analysis to identify dissolved materials, such as chemical agents and other hazardous organic compounds, is accomplished using this cell structure and the method of the present invention, in which microwave light irradiates the liquid-solute sample along the length of the cylindrical glass cell through the cell wall. The water-solute media rises in temperature during this microwave irradiation, due to absorption of the microwave energy, while concomitantly emitting increasing levels of thermal luminescence through the cell's infrared window to a sensor. The TLWM and method are pseudo-active technology in that the external microwave radiation source that heats the water sample is itself required to be outside the bandwidth of the sensor's photosensitive detector element.
More particularly, when the liquid being monitored is water, the TLWM and method embodiment of the present invention irradiates the water-solute contained in the sample cell with microwave energy of about 2.45 GHz, with 2.45 GHz being preferred. This frequency is the intense water molecular vibration-rotation absorption line, which excites the water molecules into a strong vibration-rotation state. The microwave source can be operated in pulsed or continuous-wave modes. This coupling of microwave energy into the water molecule is partly converted into heat energy. The manifestation of thermal emissions emitted by the heated sample in the 700-1400 cm
−1
middle infrared region of the electromagnetic spectrumis the detection principal of the TLWM and method. This middle infrared region of emissions is the fingerprint region where most organic contaminants of interest possess unique spectra. It is also the detection bandwidth of the system and method of the present invention.
In an exemplary embodiment of the system and method of the present invention, a Fourier Transform-Infrared (FT-IR) spectrometer is employed as a sensor device which is positioned to intercept the thermal luminescence emitted from the glass cell and to direct the thermal luminescence flux to a Michelson interferometer as its scanning component. That is, the FT-IR spectrometer collects the concomitant broad middle infrared emissions given off by the irradiated sample and then focuses it onto a photoconductive chip where the radiant thermal luminescent light is converted to interferogram temporal voltage waveforms that represent summed constructive and destructive frequencies of the collected thermal luminescence light. The interferogram is subsequently transformed, by a fast Fourier Transform operation, into an infrared Graybody spectrum that resembles a Gaussian-like profile.
Contiguous sets of these Graybody spectra are grouped as the liquid-solute sample rises in temperature, due to absorption of the irradiating microwave beam. The adjacent spectra collected when the heating rate is maximum, i.e., when
∂
2
T
∂
t
2
=
0
where
T=temperature of the sample
t=microwave irradiation time
are isolated, subtracted, and pre-processed by computer algorithms for submission to a neural network.
Pluralities of interferograms/spectra are acquired/pre-processed in this manner during a period of microwave irradiation. The system and method of the present invention thus measure a thermal luminescence difference-spectrum during an irradiation interval associated with a peak heating rate (or state of maximum thermal luminescence flux) and analyze the resulting spectral information to identify contaminants through their fingerprint absorption profiles. For this analysis, the system and method of one embodiment of the present invention employ a neural network to perform pattern recognition on the spectra of thermal luminescence collected by the sensor device. The neural network employed acts as a filter, i.e., it is trained to recognize spectral patterns o
Carrieri Arthur H.
Colclough Stephen J.
Orndoff James R.
Roese Erik S.
Schlitzkus Peter J.
Biffoni Ulysses John
Mariam Daniel
The United States of America as represented by the Secretary of
LandOfFree
Thermal luminescence liquid monitoring system and method does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Thermal luminescence liquid monitoring system and method, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Thermal luminescence liquid monitoring system and method will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3229870