KR100568891B1 - Sensor system - Google Patents

Abstract

The present invention provides a sensor system 800 for generating a sensor signal S _out corresponding to a filtered image of a scene. System 800 is more or less an array 526 and associated processing circuits for providing device signals in response to images projected on the array and projection lens set 502 for projecting images on array 526. It consists of. Lens set 502 projects images of differently blurred scenes onto array 526 to generate a measurement signal during the measurement phase and a local spatial average correction signal during the correction phase. Light diffusing shutter 522 is used to generate differently blurred images. System 800 subtracts the measurement and calibration signals of each device from the others to provide a difference signal and uses this signal to generate a sensor signal S _out during the measurement phase. Since the calibration phase is repeated at changes in the array temperature or incident radiation intensity, fixed pattern noise in the sensor signal S _out is reduced in spite of changes in the environment, rarely interrupting only the measurement phase.

Sensor system, FPN correction, scattering means, actuation means, projection means

Description

Sensor system

The present invention relates to a sensor system for sensing radiation of a scene. It also relates to a method of detecting such radiation.

Radiation sensor systems are well known in the art. They are also widely applied, for example, in thermal imagers used in portable consumer video and digital cameras and emergency services.

A typical system includes a sensor that includes a two dimensional array of elements, each having an associated signal processing circuit. Emission from the scene is projected onto the array and each element reacts with the output S _k as given in equation [1] through the processing circuit associated with it; Where index k is used to uniquely identify the elements, that is, S _k is the output of the k-th element circuit. The output S _k contains unwanted artefacts that occur in the element or processing circuit associated therewith or occur in the scene itself.

here,

S _k = output generated from the k th element through the processing circuit associated with the k th element;

A _k = k-th device reactivity function;

Rf _k = feature contrast or scene contrast emission from the scene received at the kth element;

Rped _k = background radiation from the scene received at the kth element;

B _k = offset signal generated in the k-th element and the processing circuit associated with it;

N _k = noise signal generated in the k-th element and its associated processing circuit;

Outputs S _k from each device are combined to provide a sensor signal.

The sensor including associated circuits may be based on charge coupled devices (CCDs) or metal oxide semiconductor (MOS) devices. In particular when MOS devices are used, it is found that there are unwanted deviations in the device reactivity function A _k , ie the devices have different reactivity and provide different outputs S _k in response to the same received radiation intensity. This deviation is often greater than the deviation of sensors that include CCDs. Despite long-standing pursuits to benefit from the compatibility of MOS detection and processing circuitry, the widespread use of sensors containing MOS devices has been hindered than those containing CCDs in consumer video cameras. In this regard, providing power and control signals to operate MOS devices tends to be less complex and less expensive than providing them to operate corresponding CCDs; This is because in the former case circuit parameters such as supply voltages are compatible. The deviation causes FPN (fixed pattern noise) to occur at outputs S _k , which results in a corresponding sensor signal depicting the speckled scene. Moreover, the elements also have different values of the offset signal B _k between the elements. For sensors used to respond to low levels of radiation intensity, noise N _k generated in their detector elements and associated processing circuits is often problematic, in particular flicker noise being inversely proportional to frequency. Contribute to N _k with increasing noise spectral density.

When the sensor receives the radiation from the scene, the output S _k includes an unwanted pedestal component corresponding to the general background radiation from the scene along with the characteristic or contrast component corresponding to the features in the scene. This is especially

(Iii) the sensor detects infrared radiation;

(Ii) the scene is at an ambient temperature of about 300 K; And

(Iii) when the temperature deviations within scene Rf _k are less than 1K, generating a feature component; proper.

The pedestal component may be a factor of one thousand or more greater than the feature component. This causes bad signal contrast and will make it difficult to identify temperature deviations at outputs S _k unless other signal processing is applied.

The presence of the pedestal component imposes constraints and restrictions on the design and performance of a sensor system that detects emissions from the scene, in particular infrared emissions. This often results in a sensor signal representing an image of the scene, which is governed by offset errors and artificial results with respect to scene contrast.

A solution to the problem of the pedestal component described above is described in US Pat. No. 5,155,348. This describes a readout circuit for a sensor comprising a two-dimensional array of 128 × 128 photodetector elements responsive to infrared radiation, with each device connected to a respective readout circuit. The circuit has calibration and measurement aspects.

During the calibration phase, a calibration image is projected onto the elements. The image may correspond to an uncharacteristic correction object at a temperature similar to the scene to be viewed or an overall blurry and uncharacteristic uniform image of the scene. Each device generates a signal in response to a calibration image and each circuit of the device is arranged to store a calibration signal on a storage capacitor C _c included in the circuit in response to the signal generated by each device reacting with the calibration image. do. This provides a correction for the pedestal component across the array.

During the measurement phase, a focused image of the scene is projected onto the array and generates a measurement signal at each element. The calibration signal is subtracted from the measurement signal for each device to provide a difference signal. The difference signal is integrated to provide an output signal. Circuits produce each output signal scanned by a multiplexer to provide a mixed sensor signal.

This solution reduces the dynamic range of the mixed sensor signal by eliminating the pedestal component in each device. This alleviates the requirements of the dynamic range performance of remote circuits, for example, receiving sensor signals from a multiplexer that allows the use of 8-bit resolution analog-to-digital converters instead of 12-bit resolution.

Techniques for reducing FPN are described in US Pat. No. 5 373 151. The specification describes optical projection of focused and periodically defocused images of a scene onto a multielement focal plane array for generation in each device corresponding to each of the focused and defocused signals. It's about the system. The difference signal for each element is derived by subtracting its associated defocused signal from its focused signal. The difference signals from the elements are combined together to provide a system output in which FPN artificial results are reduced.

Despite the use of prior art FPN correction, it is found in the experiment that the prior art image quality does not remain as good as the original correction suggests and needs to be recalibrated repeatedly. The reasons for this contradiction are not currently described in the prior art.

It is an object of the present invention to provide a sensor system with more effective FPN correction.

The present invention is a sensor system for generating a sensor signal corresponding to a filtered image of a scene,

(Iii) detection means including a plurality of detector elements for generating element signals during the first and second detection phases; And

(Ii) processing means for processing device signals from two detection phases, and generating processing means for generating a sensor signal from which fixed pattern noise is counteracted,

The system is characterized in that it comprises sensing means responsive to illuminance of the detection means by incident radiation affecting the detector element responsiveness and is configured to repeat both detection phases in response to a change in illuminance.

The present invention provides the advantage that the sensor system performs a recalibration to reduce FPN if necessary in response to changes in illuminance affecting the responsiveness of the detection means.

The first problem with the detection circuit described in patent specification US 5 155 348 (Ballingall and Blenkinsop) and the system described in patent specification US 5 373 151 (Eckel et al.) Is that the reactivity function A _k of equation 1 of their detector elements Is affected. As a result, the FPN correction described in these two specifications becomes inaccurate when the temperature of their detector element changes. In addition, a second problem is that the deviation of the function A _k of the detector elements will vary with the received radiation level; For example, in more or less detector arrays, the elements often exhibit the same responsiveness at relatively higher received radiation levels and different reactivity at relatively lower received radiation levels. This means that FPN correction determined at relatively higher received radiation levels will be incorrect if used at lower received radiation levels. This temperature and radiation level sensitivity of the reactivity function A _k of the devices is not identified in the prior art.

The conventional approach to solving the first problem is to stabilize the temperature of the detector array somewhat, for example, by using temperature control elements thermally coupled with the detector array. Temperature control elements may include, for example, Peltier elements, although Peltier elements have disadvantages associated with bulk, cost, high power consumption and slow thermal response. In addition, many conventional optical systems include iris components to stabilize the overall detector array illumination; These aperture components are often used in conventional 35 mm film cameras that include automatic exposure control. These aperture components are often precision optical components that are easy to mechanically mount.

Although FPN reduction is described in patent specifications US 5 155 348 (Ballingall and Blenkinsop) and US 5 373 151 (Eckel et al.), Regular and frequent recalibration to reduce FPN,

(Iii) interruption of the sensor signal occurs during the associated calibration phase; And

(Ii) Optical components that are mechanically operated to provide blurry and non-blur images are undesirable because they are associated with easy mechanical mounting and unreliable problems.

When the sensor outputs at 25 frame / second in high frame rates, e.g. video applications, the periodic interruption of its sensor signal for recalibration, for which the calibration phase is almost equal to or greater than the measurement phase in duration, Not desirable Mechanical operation of the optical components is a relatively slow progression, for example to operate a 50 g lens through a distance of 1 mm using a compact linear actuator to project focused and defocused images. It will often take 50 msec using compact linear actuators such as solenoids. Interruption of the sensor signal can be flicker, for example, to control instability problems when the sensor signal is used in a visual feedback control system. Visual feedback control systems include, for example, robotic fruit picking equipment and safety inspection systems for use in agriculture.

The above-described problems, according to the present invention, provide patent specifications for carrying out the calibration phase in response to illuminance that affects sensor characteristics that change more than a threshold amount when compared to its value when the calibration phase was most recently performed. As described in US 5 155 348 (Ballingall and Blenkinsop) and US 5 373 151 (Eckel et al.), By reducing the arrangement of sensors suitable for implementing calibration and measurement phases to reduce FPN.

In the system of the present invention;

(Iii) the sensing means also preferably reacts to a change in temperature of the detecting means;

delete

delete

delete

(Ii) The system is preferably configured to repeat the first and second phases in response to a detection means temperature change greater than a preset value.

FPN reduction in the sensor can be achieved by making one of the images a local spatial averaged image of the scene. The local spatial average image can be obtained by making one of the images a diffuse image of the scene. Therefore, the system may comprise projection means for projecting scene images on the detection means such that at least one device signal is generated during the diffuse image phase.

The projection means may comprise actuation means for inserting radiation scattering means between the scene and the detection means. This offers the advantage of a practical mechanical arrangement for generating diffuse images. For example, scattering means can be inserted to generate diffuse images with relatively low mechanical precision in comparison with accuracy, the projection means having to be maintained for the detection means for projecting the finely focused image thereon; Therefore, the system can be converted from projecting a finely focused image to projecting a blurry image with only an incorrect insertion of scattering means which does not disturb the position of the projection means relative to the detection means.

The actuation means may comprise at least one solenoid means and means for translating and rotating the scattering means for inserting the scattering means between the scene and the detection means. This provides the advantage that the actuation means can be suitably adapted to the various mechanical pressures imposed by the external packaging of the system, for example in the form of an enclosure, cabinet or case of the system.

The scattering means may comprise at least one of a ground glass plate, a translucent plastic sheet, a tracing paper, a microprism sheet, a fresnel plate and a phase plate. This offers the advantage of choosing the actual integers used to generate the diffuse images.

The projection means can be configured such that radiation corresponding to the radiation that can be received on one element from the focused image is received between the two elements and received by 64% of the elements when the image is diffused. . This offers the advantage that it is useful for practical blurry areas to achieve FPN reduction.

System

(Iii) projection means for projecting scene images onto the detection means,

(Ii) the projection means are arranged to project a local spatial average image on the detection means during one of the phases, and to project a focused measurement image during the other of the phases,

(Iii) the processing means may be configured to generate the sensor signal from the device signals derived from the local mean and measured images.

This provides the advantage that FPN reduction is achievable by including projection means into the system to project a local spatial average image of the scene onto the detection means.

The measured image can be projected onto the detection means before the local spatial image, thus providing image tone inversion in the sensor signal. This provides the advantage of a system that simultaneously performs FPN and tone inversion. Tone reversal here is defined as areas of the scene that are more radiated than other areas and are displayed as relatively dark areas in the sensor signal compared to other areas indicated in the signal.

The projection means may comprise a liquid crystal spatial light modulator configured to be able to be controlled in both cases between the first substantially non-scattered state and the second diffused state with respect to radiation transfer from the scene to the detection means. This provides the advantage of a simple approach to generating diffuse images that do not require the use of mechanically operated components that are easy to mount.

The liquid crystal modulator may be a polymer dispersed liquid crystal device (PDLC) having scattering and non-scattering states selectable in response to a control potential applied to the device. This offers the advantage of being a simple device that is conventionally switchable from one state to another and does not require polar filters to operate dissimilar types of liquid crystal devices. In addition, it is denser than diffuse shutters that respond faster and operate mechanically.

The system may be included in one or more digital still cameras, video cameras, personal electronic organizers, and mobile phones. When the system is used in digital still cameras or digital cameras, it provides FPN correction that allows the CMOS detectors used in these cameras. Personal electronic devices often operate at limited supply potential from their batteries to make use of configured electronic components; The present invention allows for example a sensor system that uses CMOS devices to configure organizers to image and scan documents shown to the organizers. The present invention constituting a mobile telephone provides a video phone by providing a data entry of images to the mobile telephone; The present invention provides the advantage of improved image quality, especially when mobile phones are rarely used and are subject to extreme temperature variations in use.
In another aspect, the invention provides a method of enhancing an image of a scene obtained by a sensor system comprising a somewhat self-detecting detector, the method comprising:
(a) detecting the radiation intensity incident on the detector;
(b) recalibrating the system to neutralize the change in fixed pattern noise in response to a change in radiation intensity incident on the detector.

The invention also provides a method of enhancing imaging by a sensor system comprising a detector having a plurality of detector elements configured to detect incident radiation intensity, the method comprising:

(a) projecting a first image onto the detector during the first phase to provide a first signal at each device;

(b) projecting a second image on the detector during the second phase to provide a second signal at each device, wherein one of the first and second images is focused and the other is partially blurred; Projection step;

(c) processing the first and second signals at each device to neutralize the fixed pattern noise and provide a difference signal, respectively; And

(d) repeating steps (a) to (c) in response to radiation intensity incident on the detector that changes to a value that exceeds a predetermined threshold.

delete

delete

delete

delete

In another aspect, the present invention solves the problem of generating a local spatial mean image that is used to perform FPN correction. In the prior art, for example, in the systems described in patent specifications US 5 155 348 (Ballingall and Blenkinsop) and US 5 373 151 (Eckel et al.), A local spatial average image is generated by projecting a defocused image of a scene. . This includes operative optical projection components between the states in which they project focused and defocused images.

In this aspect, the present invention provides a sensor system for generating a sensor signal corresponding to a filtered image of a scene,

(Iii) detection means including a plurality of detector elements for generating first and second element signals, respectively, during the first and second detection phases; And

(Ii) providing said sensor system comprising processing means for deriving a difference signal from the device signals for use in generating a sensor signal,

The system is characterized in that it comprises projection means for projecting scene images onto detector means 528 such that at least one device signal is generated during the diffuse image phase.

The present invention provides the advantage of a practical approach to generating a local spatial mean image. For example, by inserting scattering means having a relatively low mechanical precision compared to the accuracy of the projection means, a diffuse image can be generated during the diffuse image phase, the projection means being provided to a detection means for projecting a finely focused image thereon. Must be maintained in relation to it; Therefore, the system can be converted from projecting a finely focused image to projecting a blurry image with only an incorrect insertion of scattering means which does not disturb the position of the projection means relative to the detection means.

In order that the present invention may be more fully understood, embodiments thereof will now be described, by way of example only, with reference to the accompanying drawings.

1 is a schematic diagram of a sensor system.

2 is a schematic diagram of a prior art focal plan array of the system of FIG.

3 is a graph of detector signals along an axis on the array of FIG.

4 is a graph of the signals of FIG. 3 corrected by global pedestal correction and also by local pedestal correction.

5 is a schematic diagram of an embodiment of the invention that includes random access memory separated in an array.

6 is a graph showing device reactivity as a function of received radiation intensity.

7 is a graph depicting device reactivity as a function of device temperature.

8 shows an alternative embodiment of the invention of FIG. 5 including illuminance and temperature sensors.

9, 10, 11, 12 are four diagrams recorded using the embodiment of the present invention of FIG. 5 showing the image quality enhancement provided by FPN correction.

FIG. 13 is an outline view of the views shown in FIGS. 9-12 in which the positions of the FPN spots are shown in circles in the outline view.

FPN reduction will now be described with reference to FIG. 1. In FIG. 1, a schematic diagram of a sensor system, indicated generally at 1, is shown. This includes the object 2, the imaging lens 4 and the focal plane array 6. The array 6 comprises a substrate 8 to which an array of uniformly distributed detector elements, such as element 10, is glued onto its front element surface 12. Surface 12 is a rectangular region with a longer side length h as shown. Each device maintains a constant spacing between neighboring devices that are closest to each other at a distance p between the centers, as shown. The object 2, the lens 4 and the array 6 are located in order on the optical axis zz ′ perpendicularly intersecting the surface 12. The array 6 consists of three alternative positions, namely partially blurred and fully blurred uniform images of the object 2 projected onto the array 6 by the lens 4, each of which is a focused position. , Positions Q ₁ , Q ₂ , Q ₃ . At positions Q ₁ , Q ₂ , Q ₃ , the array 6 is at distances of 0, w ₁ , w ₂ along the axis zz ', respectively. The elements react to photons emitted in the corresponding regions of the object 2 to provide detector signals.

The radiation emitted and reflected from the object 2 transmitted through the lens 4 is defined as an angular range of 0 to α _max radians relative to the axis zz 'when incident on the array 6 at position Q ₁ . The radiation incident on one element 10 of the array 6 at position Q ₁ is F _max from element 10 at positions Q ₂ , Q ₃ as provided by equations [2], [3], respectively. Spread on surface 12 at a distance to.

For the purposes of this specification, the focused image is substantially in focus, ie F _max is less than the distance p. Partially blurry images may have spatial features identified but their fine spatial features are blurred. The defocused image is uniformly blurred so that the spatial features cannot be substantially identified.

In FIG. 2, a front view of the surface 12 of the array 6 of FIG. 1 seen in the zz 'direction is shown. For clarity, an array comprising a two-dimensional field of 5 by 5 elements is shown in FIG. 2 to represent an array 6 containing 128 by 128 elements, i.e., a field of 128 columns of elements and 128 rows of elements. do. The rows and columns are perpendicular to each other. The axis xx 'is shown parallel to the surface 12 and shows a row of devices, i.e., element E ₁ near one edge of the array 6 and the other edge of the array 6, as shown. Intersect elements in a row including element E ₁₂₈ . Devices E ₂ through E ₁₂₇ are located in sequence adjacently along axis xx 'between device E ₁ and device E ₁₂₈ , for example device E ₂ is located between device E ₁ and device E ₃ , and device E ₁₂₇ is positioned between the device E ₁₂₆ and the device E ₁₂₈ , and the like.

In FIG. 3, one row from device E ₁ to device E ₃₀ located from each of devices E ₁ to E ₃₀ , ie at about one quarter of the distance from device E ₁ to device E ₁₂₈ along axis xx '. From the elements of, a graph of detector signals is shown. The graph has an abscissa axis 31 representing the device position distance along axis xx '. It also has an ordinate axis 31 corresponding to detector signals from the elements. The axis 31 comprises a discontinuity line 32 in which the contrast information of the graph is exaggerated relative to the ordinate axis 31. The bends 33a (solid line), 33b (dash line), 33c (chain line) are detectors for the image focused at position Q ₁ , the image partially blurred at position Q ₂ and the defocused image at position Q ₃ . Corresponding to the signals respectively. The curvature 33a corresponds to the elements E ₃ , E ₄ , E ₁₁ , E ₁₂ , E ₁₃ , E ₁₄ , E ₁₉ , E ₂₀ corresponding to radiation emission deviations resulting from temperature deviations in the object 2. Includes feature information. The deviation of the elements E ₃ , E ₄ is smaller than the deviations of the elements E ₁₁ , E ₁₂ , E ₁₃ , E ₁₄ , E ₁₉ , E ₂₀ . The curve 33c corresponds to the average radiation emission from the object 2 and is used as a correction signal for pedestal component removal as in the prior art. Irregularities in the curvature 33c arise from differences in reactivity between the elements. The curvature 33b corresponds to the local mean dissipated emission from the entity 2 changing across the array 6.

In FIG. 4, a graph of detector signals from each of the elements E ₁ to E ₃₀ subtracted from the bend 33a to show the bends 33b and 33c respectively provide the bends 35b and 35a. Curves 35a and 35b are shown with solid and dashed lines, respectively. The abscissa axis 30 corresponds to the device position distance along axis xx 'in FIG. 2. Line 36 corresponds to a zero value. The ordinate axis 38 corresponds to the corrected device signal from which the pedestal component for each device is subtracted. The curve 35a corresponds to the relative spatial emission from the object 2 as at the radiometer output. The curve 35b corresponds to a filtered version of the curve 35a in which local deviations are emphasized upon release from the object 2.

The curve 35a is included in the first dynamic range of A _a to A _b . Local deviations in elements E ₃ and E ₄ fall within the second dynamic range of D _a to D _b . Removal of the local pedestal components results in curvature 35b falling within the third dynamic range of B _a to B _b . The fourth dynamic range of C _a to C _b includes a local deviation of the elements E ₃ , E _{4 in} the bend 35b.

The ratio of the fourth dynamic range to the third range is greater compared to the ratio of the second range and the first range.

As a result, the curvature 35b includes small local deviations of the emission from the individual 2 that do not correspond to the relative radiometer output but are highlighted in relation to the larger deviations occurring from other areas of the individual. Thus, feature edges are emphasized in the curvature 35b which increases the visual clarity of the object 2 as determined from the object.

The degree of partial blurring provided at position Q ₂ may be selected to provide the desired degree of filtering at the bend 35b. If the point of the image focused at position Q ₁ becomes 100% blurred in the image at position Q ₂ , the bends 35a and 35b are the same as when no filtering is achieved. If the blur is reduced to approach full focus, i.e. if the positions Q ₁ , Q ₂ coincide, the curvature 35b will be zero for all elements. Indeed, the degree of blurring may be chosen such that, for example, a point in the image at position Q ₁ is blurred in a range that spreads beyond the most recent consecutive elements p up to 25% of the elements of the array 6. The blurring degree can be made adjustable by manual or automatic control to provide the desired filtering degree.

To further explain the operation of the system 1, the image of the object 2 projected by the lens 4 onto the elements E arranged in the Cartesian xy plane constituting the surface 12 is expressed in [5]. Is described.

here

x ₀ , y ₀ = x-axis, y-axis Cartesian coordinates of the plane containing z = 0 at the focus, ie at position Q ₁ ;

z = z-axis Cartesian coordinates along position Q ₂ , for example;

when x, y = z is not equal to 0, for example at position Q ₂ , the x-axis, y-axis Cartesian coordinates of the plane containing the blurred image;

K = blurry function provided by lens 4; And

L = function describing the image

The curve 35b corresponds to the difference signal D (x ₀ , y ₀ ), which is the difference between the first and second images, the first image having z ordinate at position Q ₁ with z ₁ = 0, and the second The image has a z ordinate with z ₂ = w ₁ at position Q ₂ as described by equation [6].

Under the condition w ₁ = 0, the functions L (x, y, z ₁ ) and L (x, y, z ₂ ) will be the same and will contain both feature information but zero D (x, y) Will be. This happens as a result of the curvature 35b without any feature information when identically blurred images are projected onto the array 6 at positions Q ₁ , Q ₂ ; This occurs when the positions Q ₁ , Q ₂ are the same.

By controlling the function K with geometrical optical forcings, the limit F _max for the degree of blurring obtainable from the lens 4 is described by equation [7].

here

α _max = maximum radiation half-angle with respect to the axis z-z 'of radiation forming an image on the array 6.

Therefore, if the focused image is represented by an array 6 of position Q ₁ with z = 0 in equation [7], image features smaller than ztanα _max result in the curvature 35b indicated by D (x, y). Will be highlighted with filtered output. These image features will correspond to non-uniform spatial components present in the affected image by blurring. These spatial components generally correspond to characteristic information that is important for interpreting an image.

The sensor system 500 of the present invention is shown in FIG. This includes a camera 502, a controller unit 504, a frame memory (RAM) 506, an analog-to-digital converter (ADC) 508, two address latches 510, 512, a demultiplexer (DEMUX) 514. And a subtractor unit 516. They are connected together as shown in the figure and as described later.

The camera 502 is a metal oxide semiconductor (MOS) that photodetects elements 528, such as the imaging lens unit 520, the translucent shutter 522 actuated by the actuating mechanism 524, and the element 528a. Array 526. Lens unit 520, shutter 522, and array 526 are disposed along an optical axis (not shown) of system 500. The shutter 522 is mounted closer to the lens unit 520 than to the array 526; This provides a system 500 having an image that is more diffused for correction than is available when the shutter 522 is mounted closer to the array 526 than the lens unit 520. Biasing resistors, clock signal lines, and power supply lines associated with integers 520 through 528 are not included in FIG. 5 because are well known to those skilled in the art.

Array 526 is a patented FUGA-15 device manufactured by C-CAM Technologies SA that includes an array of 512 x 512 that photodetects devices that respond to radiation in the 400 nm to 1000 nm wavelength range. It includes electrical circuits to multiplex the output from each photodetector element 528 to provide an analog output S ₁ therefrom. The array also includes address inputs ADX and ADY and enabling inputs ENX and EXY. The circuits are configured to provide a signal at output S ₁ from element 528 referenced by the addresses provided at inputs ADX, ADY when the inputs ENX, ENY are set to enabled.

The controller unit 504 is configured to operate at a 40 MHz clock rate with a Max9000 of a fully programmable gate array (FPGA) type manufactured by Altera. RAM 506 is Cypress Semiconductor Ltd. It has one set of four patented static random access memory devices of type CY62170 manufactured by. It is arranged in 512 x 512 memory cells, each cell storing 16 bits of information. ADC 508 is available from Analogue Devices Inc. 12-bit pipelined semi-flash converter device type ADS800 manufactured by and a video amplifier type AD817AN manufactured by Analogue Devices Inc. to buffer between the array 526 and the semi-flash converter device. The latches 510 and 512 are each manufactured by Quality Semiconductor Ltd. Two QS74FCT373 octel latch devices fabricated by The demultiplexer 514 is comprised of Quality Semiconductor Ltd. which is configured in pairs to provide two outputs S _3A , S _3B and a single input (S ₂ ). QS74FCT541 of the four octel buffer devices types manufactured by the present invention. Subtractor unit 516 is a known type of standard parallel input, parallel output 12 bit device.

The mechanism 524 is arranged to move the shutter 522 between the first and second positions in accordance with the control input SERVO provided to the mechanism from the controller unit 504.

The shutter 522 is located in a first position so as not to intercept radiation passing from the unit 520 to the array 526 and in a second position to scatter radiation passing from the unit 520 to the array 526. do. Shutter 522 includes a ground glass plate that transmits and scatters visible radiation while diluting substantially only 5% of the radiation incidence. This provides a scattering feature in which radiation received by the element 528 in the central area of the array 526, i.e., the reference element, is scattered when the shutter is in the first position, 90% of which is in the array 526 in the second position. Illuminate with 64% of the elements 528 consecutive to the reference element in the range of elements most recently adjacent to the reference element. Shutter 522 includes many light scattering sites that help to provide a blurred image that is suitable for performing calibration of system 500.

ADC 508 is arranged to convert output S ₁ to an equivalent 12-bit digital output S ₂ . The demultiplexer 514 is configured to demultiplex its input S ₂ to its two outputs S _3A , S _3B as indicated by the control unit 504 at the control input M _en .

The RAM 506 includes address inputs ADX and ADY, an enable input R _en , and a bidirectional input / output bus BUS. It is configured to store and retrieve data in the memory cell associated with the address reference at its inputs ADX, ADY when the controller 504 points it through its R _en input. Data is input to and output from the RAM 506 via its bus.

Subtractor 516 is configured to calculate the difference between the data inputs of its inputs IN ₁ , IN ₂ to provide an output S _out from system 500.

The operation of system 500 will now be described with reference to FIG. Operation of the camera 502 begins by pointing the camera towards a scene marked 'S'.

In the calibration mode of the system 500, the mechanism is instructed by the controller unit 504 to move the shutter 522 to a second position to insert a shutter between the lens unit 520 and the array 526. The scene emits visible radiation sent by the unit 520 through the shutter 522 to form a diffused and projected image of the scene on the array 526. The elements 528 react with the diffuse image and the output of each element is multiplexed to the output S ₁ by instructions from the controller unit 504. ADC 508 converts output S ₁ from array 526 to a digital signal by its output S ₂ . Output S ₂ is demultiplexed into output S _3B which is input to RAM 506 via demultiplexer 514 and is then stored. Thus, the output from each element 528 of the array 526 is stored in a corresponding memory cell in the RAM 506 to provide calibration data.

In the measurement mode of the system 500, the camera 502 faces the scene as in the calibration mode described above. Mechanism 524 is instructed by controller unit 504 to move shutter 522 to a first position that does not block radiation from the scene. Unit 520 projects the non-diffused image of the scene onto array 526 and elements 528 respond to it. The array 526 is instructed by the controller 504 to multiplex the output from one of the elements 528 that provides an output S ₁ that is converted in the ADC 508 to a digital signal of its output S ₂ . The output S ₂ is then multiplexed through the multiplexer 514 to the input IN ₁ of the subtraction unit 516. The control unit 504 also instructs the RAM 506 to output the data stored on its bus to the input IN ₂ corresponding to the element 528 at the input IN ₁ . Subtraction unit 516 calculates the difference between the signals of inputs IN ₁ , IN ₂ and outputs this difference at output S _out to provide an output from system 500. The controller unit 504 repeats this subtraction operation for each element 528 in the array 526.

The output S _{out corresponds} to the diffuse image of the scene minus the response to the non-diffuse image of the scene to the response of each element 528. Flicker noise and deviations in device offset and reactivity, ie defects due to FPN, are reduced in the data provided at output S _out compared to the output of S ₁ directly from array 526 during measurement mode. Defects due to temperature dependence of device offset with temperature and reactivity are also reduced at output S _out .

System 500 provides the advantage of using generally available components used in the prior art to practice the present invention. Shutter 522 provides a diffuse image for FPN reduction without the requirement to change the position of lens unit 520 relative to array 526 when switching between calibration and measurement modes. In addition, the shutter 522 may be inserted between the lens unit 520 and the array 526 relatively inaccurately and with blurred images generated for FPN reduction by moving the lens unit 520 relative to the array 526. In comparison, it also provides the advantage of FPN reduction.

System 500 may be used for a plurality of applications, for example,

(Iii) in digital still cameras for recording static images;

(Ii) in video cameras for recording moving images;

(Iii) for example in personal electronic organizers for recording documents;

(Iii) in mobile telephones, may be used.

When the system 500 is used in a digital still camera or video camera, it provides FPN correction that allows CMOS detection elements to be used in these cameras. Personal electronic organizers often operate with limited supply potentials available from their batteries for the electronic components configured therein; Sensor system 500 using CMOS detection elements may, for example, configure organizers for imaging and scanning documents presented in organizers. The present invention, which constitutes a mobile telephone, provides a videophone by providing a data entry of images to the mobile telephone; The present invention provides an advantage for improved image quality.

The output S _out may be displayed on the screen and may be stored or printed in a memory device for other display or other processing, where the system 500 forms part of an electronic camera system. The memory device may be a rewritable EEPROM that provides data retention when the electrical supply is removed.

The MOS array 526 of the system 500 may be replaced with a CCD array. System 500 will provide similar benefits as in MOS array 526 in this embodiment, ie, reducing flicker noise, FPN, and offset drift. In addition, the mechanism 524 may include one or more of a motor and a solenoid for at least one of moving the shutter 522 linearly and rotationally.

In addition, the lens unit 520 may be moved by the mechanism 524 for the array 526 during the calibration mode to project an image of a partially blurred scene onto the array instead of using the shutter 522.

The shutter 522 may include a translucent plastic sheet, a tracing paper, a microprism sheet, one or more Fresnel plates, or one or more phase plates instead of a ground glass plate to diffuse radiation from the scene. Can be. The shutter 522 may be mounted as important as in a standard (SLR) reflective camera. Alternatively, it may be mounted on a rotatable carrier for quick insertion and extraction between the array 526 and the lens unit 520.

The shutter 522 and the mechanism 524 for actuating it may be replaced with a liquid crystal shutter configured to operate between transparent and partially opaque states; Liquid crystal shutters are in the form of spatial light modulators. In this case, the movement of the shutter 522 is not necessary when switching between calibration and measurement modes. Liquid crystal shutters are polymer dispersed liquid crystal devices configured to scatter visible radiation transmitted through a shutter in one state and to transmit substantially non-scattered light through the shutter in a different state depending on the bias potential applied to it. May be). PDLC contains droplets of nematic liquid crystal dispersed in a continuous isotropic polymer matrix; It is prepared by mixing monomers with liquid crystals and then polymerizing the monomers to form a solid suspension of liquid crystal droplets in the polymer matrix. The nematic liquid crystals used in PDLC may have positive dielectric anisotropy, thus scattering light therethrough when a zero bias is applied to it in the first state and substantially scattering therethrough when a bias potential is applied thereto in the second state. Transmits light. Alternatively, the liquid crystal may have negative dielectric anisotropy so that it scatters light therethrough when a bias is applied to it in the first state and is not substantially scattered therethrough when a zero bias potential is applied thereto in the second state. Transmit light. Liquid crystals with negative dielectric anisotropy comprise molecules aligned perpendicular to the electric field generated by applying a bias potential to the device; This is an advantage because the bias potential is only needed during calibration of the system 500. In addition, the system 500 may be operated to alternately run the calibration mode and the measurement mode, or may be configured to execute its measurement mode several times before returning to the calibration mode. In addition, the system may also be configured to execute a measurement mode that is first performed by the associated calibration mode when looking at the scene; This offers the advantage of being used in still cameras, where the system allows the operator to capture a scene without first running the calibration mode first.

In system 500, RAM 506 may be replaced with a dynamic random access memory (DRAM) device. DRAMs generally cost less than SRAM and thus reduce the cost of system 500. System 500 may be configured such that array 526 is replaced with an array that provides multiple signal outputs, each output connected to its own ADC whose output is multiplexed to provide signal S ₂ . This allows for slower ADCs to be used; This is particularly advantageous when the array contains more than 512 x 512 elements.

In accordance with the present invention, it has surprisingly been found that the correction required to calibrate the system 500 accurately is dependent on roughness and temperature. This is not described in the prior art, and the errors of FPN correction could be considered random in nature. However, experimental investigations for the purposes of the present invention show that the reactivity A _k varies both as a function of the radiation intensity incident on the device by the photodetecting devices 528 in the array 526 and as a function of device temperature, respectively. Indication was found. In addition, each device typically exhibits a unique reactivity A _k that will be different from that of other devices in the array 526. 6 and 7 show device responsiveness A _k as a function of device intensity and radiant intensity incident on graphs 600 and 700, respectively.

Graph 600 has an abscissa axis 602 corresponding to the radiation intensity I _R incident on the element 528, and an ordinate axis 604 corresponding to the device reactivity A _k . The three bends 610, 612, 614 represent photodetectors that react with the first, second and third devices, each having an average, higher, and lower reactivity, with respect to each other. All three bends 610, 612, 614 show the responsiveness of increasing and decreasing magnitude of radiation intensity I _R due to device reaction saturation effects.

At the radial intensity of I _R2 , the bends 612 and 614 show that the second and third elements show a difference of ΔA _A and −ΔA _{B in} the reactivity, respectively, compared to the first element of the bend 610. If the FPN correction is carried out as the prior art for radiation intensity of I _R2, to the second and third elements to -I _{R2 R2} I ΔA _A and offset correction of ΔA _B are to be matched to the first element signal in the calibration mode, Will be applied to each generated signal. However, these offset corrections are for different radiation intensities of incident I _R1 of array 526 when properly used for the second and third elements, respectively, instead of offset corrections of -I _R1 ΔA _C and I _R1 ΔA _D. Will be wrong. Therefore, the FPN correction applicable for one incident radiation intensity I _R2 will be wrong for this other intensity I _R1 because of two reasons, namely, the intensity and device reactivity are both changed.

In FIG. 7, graph 700 has an abscissa axis 702 representing device temperature T and an ordinate axis 704 representing device reactivity A _k . It has three reaction curves 710, 712, 714 for the first, second and third devices, each having an average, relatively high, relatively low reactivity. The bends 710, 712, 714 increase with increasing temperature T because

(a) the device bandgap decreases with increasing temperature T,

(b) This is because the device leakage current increases as the temperature T increases.

The reduction in the bandgap of device 528 as the temperature T increases first results in the generation of more efficient photocurrent in each device 528 resulting from the generation of photon induced electron-hole pairs; Secondly, this increases the gain in each device that converts photocurrent into an output signal. Each bend 710, 712, 714 has a different slope from each other.

At temperature T ₂ , the second and third devices have different degrees of reactivity with respect to the first device by ΔA _E and -ΔA _F, respectively. FPN correction performed according to the prior art for the temperature T ₂ and the received radiation intensity I _R is applied to the signals generated by the second and third elements to match the first element signal, respectively, -I _R ΔA Resulting in offset corrections of _E and I _R ΔA _F. However, these offset corrections will be wrong with respect to a different temperature T ₁ where offset corrections of −I _R ΔA _G and I _R ΔA _H are appropriate for each of the second and third elements. Therefore, the FPN correction determined at one temperature is wrong because the reactivity A _k changes with temperature.

When the system 500 is used in a camera, for example a still camera that records static images, it is desirable to perform both calibration and measurement modes for each scene to be observed and recorded. This is because each scene may be different in terms of illuminance compared to the previous scene. In addition, the temperature of system 500 may vary from scene to scene.

However, it is not always practical to run the calibration mode for each measurement mode when the system 500 configures a video camera because the calibration mode uses a relatively high frame update rate of about 50 frames / second. This is desirable in video cameras that make up the system 500 to implement a calibration mode that responds only to changes in illuminance or temperature. The calibration mode may be manually executed by an operator pressing a switch that configures the system 500 to instruct the controller unit 504 to perform the calibration mode.

The modified version of the system 500 shown in FIG. 8 is generally indicated at 800. System 800 is configured in system 500 except that temperature and illuminance sensors 802, 804 form an array 526 and are connected to controller unit 504 to provide temperature and illuminance data to the system. It contains the same reference items.

The controller unit 504 is configured to record data such as reference data whenever the calibration mode is executed from the sensors 802, 804. When the controller unit 504 searches for one or more current temperature and illuminance data to change above the threshold amount in the reference data, the calibration mode is executed and then the current data is recorded as the reference data. Temperature and illuminance sensors 802, 804 may be integrated or located adjacent as part of the array 526. The configuration of the temperature sensor 802 is triggerable by changes in the array 526 to allow recalibration and therefore relatively rarely auto-correction for temperature dependent FPN and offset drift while interrupting system 800 operation. to provide.

The recalibration and measurement modes described above may be used by the system 800 for a continuous time MOS detector, e.g., a FUGA-15 device, on an array 526, MOS reset detectors operating upon reset, and, for example, commercially available. It can be executed when configured with exposure cycles used in a possible JPL APS system. The continuous time detector provides an output signal from each of the elements corresponding to the radiation incident at the same time, while the reset detector provides the output signal from each of the elements corresponding to the complete radiation incident upon it during the reset time. Continuous time detectors generally exhibit larger FPN than reset detectors. When the systems 500 and 800 include a reset detector of an APS system, it is necessary to expose the diffuse image of the scene once by resetting the detector and moving the shutter 522 to the second position, and then the first position. By removing the shutter 522, it is necessary to reset and expose the scene of the non-diffused image of the scene. Continuous time detectors, such as FUGA-15 devices, offer the advantage that they do not need to be reset.

Although the systems 500 and 800 respond to visible radiation, the array 526, lens unit 520 and shutter 522 may be at different emission wavelengths, such as for example in infrared radiation and X-ray radiation wavelengths. May be replaced by the same items operable in the fields.

Referring now to FIGS. 9, 10, 11, 12, four scenes are shown as provided at the output S _out of the system 500. 9 corresponds to a scene when system 500 is calibrated and brightly illuminated. In Figure 9, FPN is not noticeable. When the illuminance of the scene is subsequently reduced to provide a dark scene without recalibration of system 500, the FPN becomes noticeable with brighter spots as shown in FIG. These brighter spots can be removed by recalibrating the system 500 for dark scenes as shown in FIG. 11. In Figure 11, FPN is not noticeable. However, when the illuminance of the scene is subsequently increased again, the FPN becomes visible with darker spots as shown in FIG. 12. Similar effects occur when the temperature of system 500 changes after calibration. Therefore, FIGS. 9-12 illustrate that recalibration of the system 500, eg, scene illuminance, is needed to maintain the image quality provided at output S _out in response to changes in environmental parameters.

FIG. 13 is an outline view of the figures shown in FIGS. 9-12, in which positions of particularly noticeable FPNs are indicated by circles in FIG.

The invention comprises one or more of the devices comprising one or more cadmium-mercury tellurium-photodiodes, photodiodes with MOS reads, phototransistors with MOS reads, photogates with MOS reads and photodiodes with CCD readouts. More or less may be used for arrays.

Although the embodiments of the present invention described above are arranged to respond to infrared and visible radiation, the present invention uses ultrasonic wave, by using a somewhat smaller array that responds to such radiation, and by using one or more focusing devices such as reflectors. It may be an alternative embodiment that is arranged in response to X-rays or microwave radiation, and is arranged for such radiation to project images of a different degree of blurring with respect to others on the array.

The systems 500, 800 may operate such that substantially focused and partially blurred images are projected onto the array 526 during the calibration and measurement modes, respectively, so that the signal corresponds to the substantially focused image. Are written into the RAM 506. This provides the advantage that image tone inversion is achieved with the signal at output S _out .

Claims (23)

A sensor system for generating a sensor signal corresponding to a filtered image of a scene, the system comprising: (a) detection means comprising a plurality of detector elements for generating element signals during the first and second detection phases; (b) processing means for deriving a difference signal from the device signals and processing the device signals from both detection phases to provide a sensor signal in which fixed pattern noise is neutralized; (c) sensing means responsive to illumination of said detection means by incident radiation affecting detector element responsiveness, The sensor system is configured to repeat two detection phases in response to this change in illuminance. The method of claim 1, (a) said sensing means is also responsive to a temperature change of said detecting means, (b) the system is configured to repeat the two detection phases in response to a change in detection means temperature greater than a preset value. The method of claim 1, And projection means for projecting scene images onto the detection means such that at least one element signal is generated during a diffuse image phase. The method of claim 3, wherein The projection means comprising actuating means for inserting radiation scattering means between the scene and the detection means. The method of claim 4, wherein The actuating means comprises at least one solenoid means and means for translating and rotating the scattering means for inserting the scattering means between the scene and the detection means. The method of claim 4, wherein The scattering means comprises at least one of a ground glass plate, a translucent plastic sheet, a tracing paper, a microprism sheet, a Fresnel plate and a phase plate. The method of claim 3, wherein The projection means is arranged such that radiation corresponding to radiation that can be received on one element from a focused image is received between two elements and received by 64% of the elements when the image is diffused. Sensor system. The method of claim 1, (a) projection means for projecting scene images onto the detection means are included in the system; (b) the projection means is configured to project a local spatial average image on the detection means during one of the phases and to project a focused measurement image during the other of the phases; (c) the processing means is configured to generate the sensor signal from the device signals derived from the local average and measured images. The method of claim 8, The projection means is configured to project the measured image onto the detection means prior to the local spatial image, whereby image tone reversal is provided in the sensor signal. The method of claim 8, The projection means comprises a liquid crystal spatial light modulator configured to be controlled in two cases between a first substantially non-scattered state and a second diffused state with respect to radiation transfer from the scene to the detection means. Which includes, the sensor system. The method of claim 10, And said liquid crystal modulator is a polymer dispersed liquid crystal (PDLC) device, said device having scattering and non-scattering states that can be selected in response to a control potential applied to it. The method of claim 1, Sensor system, which is included in one of digital still camera, video camera, personal electronic organizer and mobile phone. A method for enhancing imaging of a sensor system comprising a somewhat self-detecting detector, the method comprising: (a) generating detector element signals during the first and second detection phases, (b) processing detector element signals from both detection phases to derive a difference signal from the element signals and to provide a sensor signal in which fixed pattern noise is neutralized; (c) detecting illuminance of said detection means by incident radiation affecting detector element responsiveness and repeating two detection phases in response to a change in illuminance. A method for enhancing imaging by a sensor system comprising a detector having a plurality of detector elements configured to detect incident radiation intensity, the method comprising: (a) projecting a first image onto the detector during a first phase to provide a first signal at each device; (b) projecting a second image on the detector during a second phase to provide a second signal at each device, wherein one of the first and second images is focused and the other is partially blurred The projection step; (c) processing the first and second signals at each device to neutralize fixed pattern noise and provide respective difference signals; (d) repeating steps (a) through (c) in response to the radiation intensity incident on the detector to change and exceed a predetermined threshold. The method of claim 14, Radiation corresponding to radiation received on a detector element from the focused image is received by at least two elements and from the partially blurred image by 64% or less of the detector elements . delete delete delete delete delete delete delete delete

Images