Digital Signal Processing System for Removing DC Bias in the Output of Pyroelectric and Similar Detectors

1. Field of the Invention

The present invention relates to imaging systems. More specifically, the present invention relates to circuits and systems for processing the outputs of pyroelectric detectors used in night vision systems.

2. Description of the Related Art

Night vision systems are well known in the art. Night vision systems typically include a cryogenically cooled linear detector array with an associated cryogenic subsystem, a scanning system which moves a two dimensional image across the array, and a refractive optical system which focuses energy onto the detector. The detectors in the array either sense the heat of a body or detect low light levels.

While these systems have been used for military applications, the high cost of the scanning, cooling and optical systems associated therewith has heretofore limited the applicability of same for numerous other applications. Accordingly, there has been a need in the art for a low cost night vision system.

U. S. patent application Ser. No. 08/232,893 filed on Apr. 12, 1994 by S. H. Klapper et al. discloses and claims a low cost camera for night vision systems including a focal plane array of uncooled detectors and an optically fast, compact, optical arrangement for focusing energy from an input aperture onto the array. The array includes a plurality of pyroelectric detectors which, in the illustrative embodiment, are fabricated of barium-strontium-titanate material. Yet the characteristics and therefore the output of each detector is unique with respect to sensitivity, gain and DC (direct current) offsets by way of example. In addition, the DC offset of the detector is quite large and is typically much greater than its scene signal amplitude. Therefore, without any signal preprocessing, the detector DC offset will take up most of the dynamic range at the input of an analog-to-digital (A/D) converter. This leaves only a small portion of the A/D dynamic range for the useful detector scene signal. Thus, in order to really "see" the details of the scene signal, the A/D converter must have very high digitizing resolution. However, the higher the resolution of the A/D converter, the more expensive the A/D is, and the more bits are needed in the digital portion of the system. This results in a larger more expensive system.

The conventional technique for removing coarse DC offsets involves use of a capacitive element in series with the detector output. However, it would be prohibitively expensive to add a capacitor to each detector in an array of detectors which might include as many as 164.times.328 or 53,792 detectors.

Hence, while barium-strontium-titanate uncooled detector technology provides low cost infrared detection, conventional techniques for processing and displaying the output signals provided thereby are less than satisfactory for certain high performance, low cost applications. Thus, a need remains in the art for a system and technique for processing and displaying signals provided by uncooled barium-strontium-titanate infrared detectors.

The need in the art is addressed by the present invention which, in a most general sense, provides a digital signal processing system for removing DC offset in the output of pyroelectric and similar detectors. The detector provides a first output during a first time interval and a second output during a second time interval. The inventive signal processing system including a first circuit for inverting the first output signal, a second circuit for storing the inverted first output signal, and a third circuit for adding the stored inverted first output signal to the second output signal to provide a offset corrected output signal.

In a more specific embodiment, the invention provides a digital signal processing system for removing DC offset in the output of an array of pyroelectric detectors. In addition to eliminating the DC offset in the detector outputs, the specific embodiment further includes a fourth circuit for correcting for DC offset in the signals generated by each detector in the array and a fifth circuit for correcting for gain in the signals generated by each detector in the array. In addition, a sixth circuit is included for providing automatic global gain control across the array and a seventh circuit is included for providing automatic level control across the array.

The invention takes advantage of the scene to source and source to scene gain inversion property of the pyroelectric and similar types of detectors when the detector scene switches between scene and source. The scene switching of the detector is obtained by using an optical chopper. When the scene of the pyroelectric detector switches between source (a uniform temperature source) and scene (actual target), the average DC offset of the detector remains the same, but the target signal is inverted.

The invention also removes any DC offsets introduced into the detector signal by electronic circuits (e.g. readouts, amplifiers, drivers) between the detector and the invention. It also removes offsets introduced by cabling. In the past, expensive components had to be used in order to avoid offsets. Otherwise, the offsets could be amplified by the large system gain. The invention allows less expensive components to be used.

FIG. 1 is a block diagram of the signal processing system of the present invention.

FIG. 2a is a block diagram of the analog input section of the signal processing system of the present invention.

FIG. 2b is a simplified block diagram of the analog input section and the digital signal processing section.

FIG. 3 is a block diagram of an illustrative implementation of a first digital-to-analog converter of the analog section of the signal processing system of the present invention.

FIG. 4 is a block diagram of an illustrative implementation of a first integrator circuit of the analog section of the signal processing system of the present invention.

FIG. 4(a) is a waveform illustrating the output of a typical pyroelectric detector.

FIG. 4(b) is a waveform illustrating a the waveform of FIG. 4(a) after field-to-field gain inversion in accordance with the present teachings.

FIG. 4(c) illustrates a previous field of output from a pyroelectric detector.

FIG. 4(d) is a waveform illustrating the output after a summation of the waveforms shown in FIGS. 4(b) and 4(c).

FIG. 5 is a block diagram of an illustrative implementation of a second digital-to-analog converter of the analog section of the signal processing system of the present invention.

FIG. 6 is a block diagram of an illustrative implementation of the analog output section of the signal processing system of the present invention.

Illustrative embodiments and exemplary applications will now be described with reference to the accompanying drawings to disclose the advantageous teachings of the present invention.

While the present invention is described herein with reference to illustrative embodiments for particular applications, it should be understood that the invention is not limited thereto. Those having ordinary skill in the art and access to the teachings provided herein will recognize additional modifications, applications, and embodiments within the scope thereof and fields in which the present invention would be of significant utility.

FIG. 1 is a block diagram of the signal processing system 100 of the present invention. The thermal infrared detector 20 is shown in phantom. The detector can be a focal plane array of uncooled pyroelectric bariumstrontium-titanate (BST) detectors such as that sold by the Texas Instruments company. This detector converts infrared energy into an electrical signal without the aid of an expensive cryogenic cooler as is typically required in conventional high-performance night vision imaging systems.

The characteristics of the detectors utilized in the illustrative embodiment are shown in Table I below:

In the table, "NET" refers to noise equivalent temperature and the axial and off-axis modulation transfer functions (MTF) are in units of lines per millimeter. See also U.S. Pat. No. 4,080,532 entitled FERROELECTRIC IMAGING SYSTEM, issued Mar. 21, 1978 to G. S. Hopper; U.S. Pat. No. 5,010,251 entitled RADIATION DETECTOR ARRAY USING RADIATION SENSITIVE BRIDGES, issued Apr. 23, 1991 to J. Grinberg et al.; U.S. Pat. No. 5,021,663 entitled INFRARED DETECTOR, issued Jun. 4, 1991 to L. J. Hornbeck; U.S. Pat. No. 5,034,608 entitled INFRARED SENSOR OPERABLE WITHOUT COOLING, issued Jul. 23, 1991 to L. S. Tavrow et al.; and U.S. Pat. No. 5,288,649 entitled METHOD FOR FORMING UNCOOLED INFRARED DETECTOR, to W. Koestas (Ser. No. 951,928).

Each detector has an amplifier (not shown) in a semiconductor circuit bonded to the back of the array. The amplifier provides an output signal indicative of a change in thermal energy incident on the associated detector. The amplifier outputs are scanned by a multiplexer which provides a serial analog output signal.

BST detectors are most sensitive to thermal contrasts modulated at a predetermined range of frequencies. This modulation of the scene, called "chopping" is accomplished by rotating an optical diffusing or opaque element in front of the detector at the field rate to change the detector output on every field. This provides a thermal differential to each element in the array during each revolution. The chopping momentarily focuses, then defocuses, the thermal scene from the detector element, allowing for a comparison of the scene with the average value of the scene. The result is a first output from the detector-amplifier pair which represents the detection of energy from a focused scene and a second output which is a signal of equal and opposite polarity which represents the output of the detector amplifier resulting from the detector seeing the diffused scene. These two fields constitute a single frame.

The signal processing circuit 100 of FIG. 1 receives these signals from the array 20 and corrects each pixel for sensitivity and offset. Then it offsets and scales the video signal to match the dynamic range of an analog-to-digital converter which converts the video signal to digital form. The signal processing circuit 100 includes an analog input section 200, a digital logic section 300, a timing circuit 320 and an analog output section 400.

FIG. 2a is a block diagram of the analog input section 200. FIG. 2b is a simplified block diagram of the analog input section and the digital section 300. Serial video signals are received from the infrared detector array 20. As mentioned above, the video signals consist of the two fields of alternating polarity separated by inactive intervals and offset by a DC level. This signal is passed through a buffer amplifier 210 and applied to a first adder circuit 212. The adder 212 has three inputs: the video signal, an individual pixel offset correction signal from a first digital-to-analog converter (DAC) 220, and a feedback signal from a first integrator circuit 230.

In accordance with the present teachings, in a calibration mode, a uniform reference pattern is used to illuminate the array while the outputs are scanned into the field memory 308 of the digital logic circuit 300. In the preferred embodiment, the inventive digital logic section 300 is implemented with a field programmable gate array with firmware stored in a programmable read-only memory (PROM). However, those skilled in the art will appreciate that the invention may be implemented in software in a microprocessor or in hard-wired digital logic without departing from the scope of the present invention. The ideal output of each detector should be zero but in reality it is not. As discussed more fully below, correction signals for each detector are stored in a detector coarse DC offset correction memory 302 in the digital logic circuit 300. These signals are used for offset correction of the instantaneous signals output by each detector via a DAC 220 and an adder 212, so that the detector outputs are zero at the output of the adder (212).

FIG. 3 is a block diagram of an illustrative implementation the DAC 220. A source of reference potential 222 provides a reference signal to a digital potentiometer 224. Since the range of offset correction required varies from detector to detector, the range of the digital-to-analog converter 226 is matched to each detector by the digital potentiometer 224, whereupon, the digital-to-analog converter 226 converts the stored offset signals to an analog output signal for input to the adder 212.

The integrator 230 averages two successive fields and eliminates the average offsets. The first integrator 230 cancels the DC offset of the input signal and performs a function similar to AC coupling with a long time constant. That is, the average value of the active period is set to zero by integrating the video and feeding the result back into the first adder circuit 212.

FIG. 4 is a block diagram of an illustrative implementation of the first integrator circuit 230. The switch 232 receives a blanking signal from a timing circuit 320. The RS-170 display controller timing circuit 320 provides timing for the system suitable for the RS-170 video format. The blanking signal serves to disable the integrator 234 during inactive periods in the video signal.

Returning to FIG. 2b, the output of the adder 212 is input to a unity gain inversion circuit or polarity switch 240 which operates under control of the RS-170 timing circuit 320. The polarity switch 240 is implemented with analog switches. The input/output waveform of the polarity switch 240 is shown in FIGS. 4(a)-4(b). As discussed more fully below, a significant feature of the present invention derives from the inversion of alternating fields of the output of each detector by the unity gain inversion circuit 240. This facilitates field to field subtraction which effectively cancels DC offset in each field while preserving the AC signal for subsequent processing.

FIG. 4(a) is a waveform illustrating the output of a typical pyroelectric detector. FIG. 4(b) is a waveform illustrating a the waveform of FIG. 4(a) after field-to-field gain inversion in accordance with the present teachings. FIG. 4(c) illustrates a previous field of output from a pyroelectric detector. FIG. 4(d) is a waveform illustrating the output after a summation of the waveforms shown in FIGS. 4(b) and 4(c).

As mentioned above, the chopper 19 chops the scene at the field rate to change the detector output every field. See FIG. 4(a). The polarity switch 240 effects field-to-field gain inversion in such a way that for the even fields (n=0, 2, 4, 6, . . . ), the resulting output equals (1) times the detector output and for the odd fields (n=1, 3, 5, . . . ), the resulting output equals (-1) times the detector output. See FIG. 4(b). Then the inverted signal is digitized by the analog-to-digital converter 294. Alternating fields are stored in a field memory 308. Data from a previous field (field n-1) (FIG. 4(c)) is retrieved from the field memory 308 and summed with the current field (field n) to cancel the DC bias in each field (FIG. 4(d)). As shown in FIG. 4(d), after summing fields n-1 and n together in the adder 310, the detector DC bias is canceled, but the resulting detector scene signal is doubled. Hence, a divide by two circuit 312 is used to scale the detector scene signal back to its correct signal amplitude.

A second step in the calibration process involves the illumination of the array 20 with a uniform warm reference pattern. The outputs of the detectors should be equal. The equal and nonequal outputs are digitized, computed as a function of the difference between the two reference source readings and stored in a gain correction memory 304 in the digital logic section 300 and used for sensitivity correction via a multiplier 250 and a second DAC circuit 260.

FIG. 5 is a block diagram of an illustrative implementation of the DAC circuit 260. The integrator 262 integrates pulses from the digital logic circuit 300 to create a voltage level proportional to the desired video gain. This controls the global gain applied to the detector with the weakest signal output. This gain factor is further reduced to a minimum value for the detector with the strongest output in sensitivity calibration. The range of gain control is set by a digital potentiometer 266. The integrator 262 and the potentiometer 266 feed a subtractor circuit 264 which in turn provides an input signal to a multiplying D/A (digital-to-analog) converter 265 which sets intermediate levels. As the range of the D/A converter is reduced, a complementary DC level is added to the output by an inverter 268 and a subtractor 267 to keep the full scale output of the multiplier 250 constant.

Returning to FIG. 2b, in addition to a limiter 290, two feedback loops are used to control the dynamic range of the signal going into the A/D (analog-to-digital) converter 294. The output of the A/D converter is fed to a field-to-field automatic global gain and level control circuit 306 in the digital logic circuit 300. The field-to-field automatic global gain and level control circuit 306 compares the digitized signal to upper and lower threshold values. If the digital value is above the upper threshold, a high level is output to an offset integrator 280. If the value is below the lower threshold, a low level is sent to the integrator 280. Levels between the thresholds tri-state the output to the integrator resulting in no signal from the digital logic circuit 300. This output is also tri-stated during inactive video periods. The integrator 280 is referenced to a voltage midway between the two logic levels. The result is that equal numbers of pixels will fall above the upper threshold and below the lower threshold. This provides automatic global level control across the array 20.

Automatic gain control may be effected in accordance with any one of a number of schemes are well known in the art. In the preferred embodiment, automatic gain control is provided by generating a high level when the signal is outside the previously described thresholds, a low level when it is within the thresholds, and tri-state when the video is inactive. This is fed to the integrator 262 of FIG. 5 and compared to a reference that is a fraction of the logic level (10% for example). The result is that most of the pixels (90%) will fall between the thresholds and a few pixels (10%) will be above or below the thresholds.

Thus, as depicted in FIG. 1, the digital signals from the analog input section 200 are processed by digital logic circuit 300 and stored in its frame memory. The digital output from the digital logic 300 is converted to standard video format (typically RS-170) by an analog output section 400.

FIG. 6 is a block diagram of an illustrative implementation of the analog output section 400. The video signal comes from the digital logic section 300 as successive pixel values. These digital signals are converted to analog form by a DAC 401. The scale factor of these signals is varied by using the output of an integrator 406 as the reference voltage for the DAC 401. A second integrator 407 is used to supply an offset voltage which is added to the DAC output in an adder stage 402. The output of the adder is then amplified in an amplifier stage 403 to the final output level desired. The video signal is then combined with sync and blanking pulses from the digital logic using a clipper circuit 404. The clipper circuit 404 is arranged such that the most negative signal passes to its output. The sync signal is adjusted such that its negative excursion is below the blanking and video signal levels and its positive excursion is above the video signal range. The blanking signal is adjusted such that its negative level is below the video range but above the negative sync level and its positive level is above the video signal range. When the sync signal is negative, the output is at the sync level regardless of the blanking and video signal levels. When the sync signal is positive and the blanking level is negative, the output is at the blanking level regardless of the video level. When both the sync and blanking signals are positive, the output follows the video signal. A buffer amplifier 405 provides the output current necessary to drive the load (typically 7.5 ohms) placed on the circuit. Feedback may be applied around the amplifier clipper and buffer stages to control their overall gain.

The comparators 408 and 409 are used to detect when the output exceeds the middle of video range (408) or falls outside two thresholds set near the limits of the video range 409.

The automatic level control operates by using the digital logic to gate the output of the midpoint comparator 408 with the blanking pulse. The output of the digital logic is open circuited (tri-state) during the blanking periods. It is low when the video is below the midpoint value and high if the video is above the midpoint value. This signal is integrated by an integrator 407 which is referenced to a voltage midway between the high and low logic levels. The output of the integrator controls the offset added to the video signal. An equilibrium condition is established where half of the pixels are above midscale and half are below.

The other comparator 408 is a window comparator with limits set near the limits of the video range (typically at 10% and 90% of the video range). The automatic gain control operates by using digital logic to gate the window comparator with the blanking pulse. The output of the digital logic is open circuited (tri-state) during the blanking periods. It is low when the video signal is within the window and high when the video signal is outside the window. This signal is integrated by an integrator 406 which is referenced to a voltage that is a fraction of the digital logic voltage (typically 5%). The output of the integrator varies the reference voltage of the DAC 401. This forms a feedback loop that reaches an equilibrium condition when 95% of the pixels fall between the 10% and 90% levels of the video range.

While these circuits may be implemented in a manner well within the ability of one of ordinary skill in the art, in the preferred embodiment, the circuits disclosed and claimed in U.S. patent application Ser. No. 08/226,796, filed herewith, by Masarik et al. and entitled SIGNAL PROCESSOR FOR INFRARED CAMERA the teachings of which are incorporated herein by reference, would be used for the illustrative application.

Thus, the present invention has been described herein with reference to a particular embodiment for a particular application. Those having ordinary skill in the art and access to the present teachings will recognize additional modifications applications and embodiments within the scope thereof. For example, the invention is not limited to use with BST pyroelectric detectors. The invention may be used with cooled and uncooled detectors alike.

It is therefore intended by the appended claims to cover any and all such applications, modifications and embodiments within the scope of the present invention.

Accordingly,