Automatic Responsivity Compensator in an IR Imaging System

The general field of technology of the present invention is in signal processing of multiplexed far infrared (IR) imaging systems.

One approach to far infrared imaging systems is to use a linear array of detectors coupled through buffer amplifiers to a time division multiplexer to produce a serial output corresponding to scanning the detectors along the array. A combination of optical scan in one direction and electronic scan in the other produces raster scene scan. A cathode ray tube readout may then be coupled and synchronized to produce a visible raster picture.

Recently, time division multiplexers have been constructed using charge coupled devices (CCDs). These devices can operate at low temperatures making multiplexing inside the detector dewar possible. However, the response of an array of detectors to a given input signal varies from detector to detector. This response variance requires that the gain associated with each channel be changed to normalize the outputs prior to multiplexing. This is usually done in the buffer amplifiers. In systems where the multiplexer must operate with fixed gain inputs such as a monolithic CCD device in the dewar, it is not possible to set each gain individually nor to perform automatic responsivity equalization. To make multiplexing in the dewar practical some means of equalizing the serial output must be used. The present invention accomplishes equalization of the serial output.

The present invention overcomes the disadvantages of prior art IR imagers by providing an automatic gain adjustment for each detector channel to equalize responsivity. The present automatic responsivity compensator operates on the serial output from the multiplexer and therefor is achieved outside the cryogenic dewar. Since the operation is automatic, individual controls for each detector channel are not needed and drift may be compensated automatically on each scan.

The automatic responsivity compensator is connected at the output of the multiplexer for equalizing the output voltages therefrom. The compensator is comprised of a gain amplifier that provides an active signal therefrom and a divider having as one input a signal reference which corresponds to the desired overall output signal level of the gain amplifier for a nominal responsivity of the detector array and as a second input the active signal from the gain amplifier. The signal reference is always divided by the active signal in the divider. If there is a difference between the signal reference and the active output signal, a correction signal close to unity will be generated from the divider. Otherwise, if both are equal, a unity correction signal is generated. It should be noted that there is an established correction signal of some voltage value for each of all the detector channels that are multiplexed out into the gain amplifier. If the correction signal is other than unity, the correction signal will be on the opposite side of unity than what the active signal is from the signal reference, i.e. if the active signal is higher than the signal reference, the correction signal will be less than unity, but if the active signal is less than the signal reference, the correction signal will be above unity. The simple mathematical explanation is that the division always has the dividend as a constant, representing the constant signal reference, with the divisor as a variable, representing the variable active signal. Thus, if the active signal is more than the signal reference, a division would result in a correction signal less than unity, or vice versa.

Each of the established correction signals that are generated by the divider are serially gated into a CCD serial analog memory and are stored therein. The correction signals are sequenced out of the CCD serial analog memory to either directly control the gain bias of a controllable gain amplifier in one embodiment of this invention, or as in another embodiment to indirectly operate on the output signal from a constant gain amplifier by being multiplied with the active output signal from the constant gain amplifier in an analog multiplier.

The established correction signal for each detector channel are also constantly sequenced out of the CCD serial analog memory back to its input for recirculation therethrough. The established correction signal for each detector channel operates on the gain amplifier to provide automatic gain control to normalize subsequent readout of their respective detector channels.

In reference to far infrared detectors, such as mercury-cadmium-telluride (HgCdTe) or lead-tin-telluride (PbSnTe), the output is not commensurate with the input thereto, with variation therebetween of up to 20%. The present automatic responsivity compensator functions to equalize the output voltages from such detectors. In the IR imaging systems in which the present automatic responsivity compensator is used, during scan dead time the detectors look at a fixed temperature source and the multiplexer scans all of the detectors sequentially. If the output voltage varies from the ideal as represented by the fixed temperature source, correction signals are generated out of the divider as discussed herein above. Each IR imaging system is first laboratory tested for the sensitivity and temperature responsivity of the detectors to determine what value the active signals, or output signal, from the gain amplifier should be for a nominal signal therefrom. The signal reference should then be set at this nominal value to maintain the automatic responsivity.

The invention will be better understood with reference to the detailed description herein below with reference to the following drawings.

FIG. 1 illustrates in block diagram form one embodiment of the present invention; and

FIG. 2 illustrates a second embodiment of the present invention.

Referring to FIG. 1, a detector array 10 is shown having seven representative linear detector channels therefrom which are scanned, or multiplexed, by a charge coupled device (CCD) time division multiplexer 12. The detector array 10 and multiplexer 12 are both enclosed within a cryogenic dewar (not shown).

In this type IR imaging system where the multiplexer must operate with fixed gain inputs in accordance with a fixed temperature source, such as a fixed temperature bar or the case of the IR imager, that the detectors are exposed to at the end of each scan, some means of equalizing the serial output from the multiplexer 12 is needed. The present automatic responsivity compensator provides automatic gain adjustment for each detector channel to equalize responsivity. Automatic responsivity compensator conveniently operates outside the cryogenic dewar.

The detectors of the detector array 10 have their field of view scanned by some optical means (not shown since that is not a part of the present invention). Each detector produces electrical signals therefrom according to the thermal changes in the scene as viewed by the detectors, and also due to the detectors responsivity. Multiplexer 12 electrically scans the signal inputs, from the detector channels, represented by numerals 1 through 7, according to clock signals from a controller 60, and produces serial output signals of the vertical linear detector channels.

The serial channel output signals are initially amplified by controllable gain amplifier 30. The output signal, herein referred to as the active signal, from amplifier 30 is applied to one input of a divider 26 within the automatic responsivity compensator 70. A signal reference 28, having a voltage value corresponding to the normalized output voltage V.sub.out, is applied to a second input to divider 26. If there is a difference between the active, or original, signal and the signal reference 27, the divider 26 produces a correction signal at its output. A controller 60 that controls the overall operation of the automatic responsivity compensator gates the correction signal through the closed side of a single-pole, double-throw switch of first signal gate 22, into the CCD serial analog memory 40. While a complete series of correction signals E.sub.1 through E.sub.7 is stored in memory, second signal gate 20 has a single-pole, double-throw switch therein that connects a fixed gain reference to the controllable gain amplifier 30 so that gain amplifier 30 continues to operate in an originally biased condition for fixed gain. Controller 60 also controls the operation of gate 20 along with the CCD multiplexer 12 and memory 40 simultaneously so that all the channels 1 through 7 of the linear detectors are clocked out of the multiplexer 12 and the corresponding correction signals E.sub.1 through E.sub.7 for each channel are clocked into memory 40 in a sequential manner wherein the correction signals are serially stored in the memory 40. Look at FIG. 1 more closely for an explanation of the operation of the automatic responsivity compensator after correction signals E.sub.1 through E.sub.7 have been stored in memory 40. Along the memory 40 block is shown the symbolic stored correction signals, represented as E.sub.2 through E.sub.7, with correction signals E.sub.1 shown as being fed out of memory 40 back through first signal gate 22 into memory 40. It should be noted here that after E.sub.1 through E.sub.7 correction signals have been stored in memory 40, the single-pole, double-throw switch of first signal gate 22 isolates the output of divider 26 from memory 40 so that correction signals are no longer fed into memory 40. The established signals E.sub.1 through E.sub.7 are however constantly recirculated through memory 40. While E.sub.1 is being reentered in memory 40 through gate 22, the single-pole, double-throw switch in second signal gate 20 is closed by controller 60 and signal E.sub.1 is applied to the base of a gain control circuit 32. Circuit 32 may be a field-effect-transistor (FET) of either the depletion or enhancement modes, and is preferably a normally partially conductive depletion mode n-type channel, but may be p-type as well. The above sequence of circulating the E.sub.1 correction signal is taking place while channel 1 is being multiplexed into the controllable gain amplifier 30. Any gain control that is needed because of the previously stored readout signal intensity from channel 1 being different from the signal reference 28 is applied to amplifier 30 to return the output voltage V.sub.out to a normalized value equal to the signal reference 28. Specifically, if the immediately previous readout signal intensity from channel 1 was higher than the signal reference 28 then subsequent readouts from channel 1 would be lowered in accordance with the active signal herein divided into signal reference 28, which is at a constant value that is lower, thus obtaining a correction signal less than unity that is applied to the base of the FET of gain control 32. In other words, the gain of 30 is operated on in accordance with the analog voltage level in memory 40 as established by the generated correction signals. Controllable gain amplifier 30 and gain control 32 form an analog multiplier circuit 33.

The above sequence is continuous with serially switched signals E.sub.1 through E.sub.7, representative of the total number of detector channels operated on, being constantly recirculated through memory 40 and applied to the base of the FET in circuit 32. The readout of correction signals E.sub.1 through E.sub.7 from memory 40 are synchronized by controller 60 to be applied respectively to the base of the FET at the exact time that the multiplexer is being controlled by controller 60 to read out the corresponding detector channels 1 through 7. The FET may be a junction gate FET (JFET), which is always a depletion type, or an insulated gate FET (IGFET), which may be a depletion type. The FET maintains the output voltage V.sub.out equivalent to the input from the multiplexer and at a nominal value determined in the laboratory for signal reference 28. The FET being partially conductive and controlled by the voltage applied to its base functions as a variable resistor, i.e. it is originally biased to a certain amount of conduction and varies around that amount by the voltage applied to its base. Memory 40 thus circulates each channel scan line correction signal therethrough, by controller 60 operation, to be ready for controlling the responsibility of the next corresponding scan line. After initial storage of all the sampled scan lines, recirculation is accomplished by opening one side of the single pole, double throw switch in the first signal gate 22 for isolating the divider 26 but having the other side of the single pole, double throw switch remaining closed in gate 22 to pass the recirculated correction signals back into memory 40. It should be noted that the analog memory bits must remain in memory 40 for one frame time. This is a long time by present CCD standards.

FIG. 2 illustrates an alternate type of automatic responsivity compensator 71. In this embodiment, the analog correction signal directly corrects the output from a constant gain amplifier 31 by multiplying this output with the correction signal generated by divider 26 in analog multiplier 50. The constant gain amplifier 31 has an active output signal therefrom that is applied to the first input of the divider 26, the same as in the embodiment shown by FIG. 1, but also applies the active signal to one input of the analog multiplier 50 for multiplication with the corresponding correction signal from memory 40 for the same channel. During the time that one of the correction signals, say for example correction signal E.sub.1, is being recirculated and reentered back into memory 40, the correction signal E.sub.1, is also applied as the second input to 50 and is being multiplied with the active signal from detector channel 1 out of the constant gain amplifier 31. Analog multiplier 50 operates to lower an active signal from amplifier 31 that is too high or to raise an active signal from amplifier 31 that is too low since divider 26 produces a correction signal that is on the opposite side of unity in accordance to whether the active signal is above or below the laboratory determined amount of signal reference 28. In this embodiment, controller 60 synchronizes the multiplexing out of each of the particular detector channels from multiplexer 12 and through constant gain amplifier 31 to one input of analog multiplier 50 to the application of the established correction signal generated by the same detector channel on the memory storage scan at a second input to the analog multiplier 50.