METHOD FOR PROCESSING A RAW IMAGE COLLECTED BY A BOLOMETER DETECTOR AND ASSOCIATED DEVICE

Abstract

A method for processing a raw image characterized by first Pix.sub.1(i,j) and second Pix.sub.2(i,j) raw measurements collected by first Bol.sub.1(i,j) and second Bol.sub.2(i,j) bolometers of a set of bolometers Bol(i,j) of a detector, the first bolometers Bol.sub.1(i,j) being closed off, the method being executed by a computer on the basis of reference measurements Pix.sub.REF(i,j) that include first Pix.sub.1REF(i,j) and second Pix.sub.2REF(i,j) reference measurements associated with the first Bol.sub.1(i,j) and with the second Bol.sub.2(i,j) bolometers, the method including: a) a correlation step between the first raw measurements Pix.sub.1(i,j) and the first reference measurements Pix.sub.1REF(i,j); and b) a step of correcting the raw image, which includes computing corrected measurements Pix.sub.Cor(i,j) of a corrected image for each bolometer Bol(i,j) on the basis of the reference measurements Pix.sub.REF(i,j) and of the result of step a).

Claims

1. A method for processing a raw image characterized by raw measurements Pix(i,j) that comprise first Pix.sub.1(i,j) and second Pix.sub.2(i,j) raw measurements collected, respectively, by first Bol.sub.1(i,j) and second Bol.sub.2(i,j) bolometers of a set of bolometers Bol(i,j) of a detector that are arranged in the form of a matrix-array of n rows (L.sub.i) and m columns (C.sub.i), the first bolometers Bol.sub.1(i,j) being closed off during the acquisition of the raw measurements Pix(i,j), the method being executed by a computer on the basis of reference measurements Pix.sub.REF(i,j) that comprise first Pix.sub.1REF(i,j) and second Pix.sub.2REF(i,j) reference measurements associated, respectively, with the first Bol.sub.1(i,j) and with the second Bol.sub.2(i,j) bolometers, the reference measurements Pix.sub.REF(i,j) being obtained, during a step of calibrating the detector, by masking all of the bolometers with a shutter, the shutter being kept at a reference temperature, the method comprising: a) a correlation step comprising a linear regression for each first bolometer Bol.sub.1(i,j), between the first raw measurements Pix.sub.1(i,j) and the first reference measurements Pix.sub.1REF(i,j), satisfying the following relationship:
Pix.sub.1(i,j)=α.Math.Pix.sub.1REF(i,j)+Pix.sub.offset where α and Pix.sub.offset are the terms determined during the linear regression; and b) a step of correcting the raw image, which comprises computing corrected measurements Pix.sub.Cor(i,j) of a corrected image for at least each second bolometer Bol(i,j) on the basis of the reference measurements Pix.sub.REF(i, j) and of the result of correlation step a).

2. The method according to claim 1, wherein correction step b) comprises computing the corrected measurements Pix.sub.Cor(i,j) for at least each second bolometer Bol(i,j) in accordance with the following relationship:
Pix.sub.Cor(i,j)=Pix(i,j)−α.Math.Pix.sub.REF(i,j)−Pix.sub.offset

3. The method according to claim 1, wherein the reference measurements Pix.sub.REF(i,j) are saved in a memory space of the detector.

4. The method according to claim 1, wherein the detector is provided with a lens mounted on a diaphragm that closes off the first bolometers Bol.sub.1(i,j) at the corners of the detector.

5. The method according to claim 1, wherein the reference temperature is a temperature equal to that of the detector.

6. The method according to claim 5, wherein the reference measurements Pix.sub.REF(i,j) are average measurements of the signal actually collected by the bolometers during the calibration step.

7. The method according to claim 1, wherein the detector also comprises blind bolometers Bol.sub.blind(j), each blind bolometer Bol.sub.blind(i,j) being implemented for the differential measurement of the bolometers of at least one column of bolometers Bol(i,j) that is specific thereto; each blind bolometer Bol.sub.blind(i,j) is advantageously associated with a single column (CO of bolometers (Bol(i,j)).

8. A non-transitory computer-readable medium having a computer-executable computer program stored thereon that, when executed by a computer, causes the method according to claim 1 to be implemented.

9. A device comprising: a detector provided with bolometers Bol(i,j) arranged in the form of a matrix-array of n rows (L.sub.i) and m columns (C.sub.i), the bolometers comprising first Bol.sub.1(i,j) and second Bol.sub.2(i,j) bolometers, the first bolometers Bol.sub.1(i,j) being closed off; and a computer provided with a computer program that when executed by the computer causes the method according to claim 1 to be implemented.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Other features and advantages will become apparent in the following description of a method for processing an image, given by way of non-limiting example and with reference to the appended drawings, in which:

[0051] FIG. 1 is a raw image of a scene obtained by a detector provided with bolometers arranged in a matrix-array of 80 rows by 80 columns;

[0052] FIG. 2 is a schematic depiction of a detector provided with a lens mounted on a diaphragm able to be implemented according to the present invention;

[0053] FIG. 3 is a graph-based depiction of the first raw measurements Pix.sub.1(i,j) (on the vertical axis) as a function of the first reference measurements Pix.sub.1REF(i,j) (horizontal axis);

[0054] FIG. 4a is a raw image formed by the raw measurements Pix(i,j) collected by the detector according to the present invention;

[0055] FIG. 4b is a corrected image formed by the corrected measurements Pix.sub.Cor(i,j) according to the terms of the present invention.

DETAILED DESCRIPTION OF PARTICULAR EMBODIMENTS

[0056] The present invention will now be described with reference to FIGS. 1 to 4b.

[0057] FIG. 2 shows a detector 1 provided with a plurality of bolometers, denoted Bol(i,j), arranged in the form of a matrix-array of n rows (denoted “L.sub.i”) and m columns (denoted “C.sub.j”).

[0058] A bolometer indexed i, j corresponds to a bolometer arranged at the intersection of the row i with the column j.

[0059] Among the bolometers Bol(i,j), it is possible to distinguish between first bolometers Bol.sub.1(i,j) and second bolometers Bol.sub.2(i,j).

[0060] In particular, the bolometers Bol(i,j) are arranged such that the first bolometers Bol.sub.1(i,j) are closed off while the second bolometers Bol.sub.2(i,j) are exposed to a scene intended to be imaged.

[0061] In other words, only the second bolometers Bol.sub.2(i,j) are sensitive to the (thermal) radiation from the scene intended to be imaged, while the first bolometers Bol.sub.1(i,j) do not receive or perceive any radiation from said scene.

[0062] Therefore, the definition of a first bolometer Bol.sub.1(i,j), in accordance with the terms of the present invention, may involve selecting said first bolometers Bol.sub.1(i,j) from among all of the bolometers Bol(i,j) of the detector 1. In this respect, this selection may be based on considerations in relation to the effectiveness of the closing off of said first bolometers Bol.sub.1(i,j).

[0063] It is also understood that the closing off of the first bolometers Bol.sub.1(i,j) is permanent, and that this may be achieved via a specific fixed element (for example a diaphragm as described in the remainder of the description) forming an integral part of the detector. “Fixed element” is understood to mean that said element is not mobile and not motorized.

[0064] The first bolometers Bol.sub.1(i,j) may for example be closed off by a diaphragm on which there is mounted a lens 2 that is arranged between the scene and the detector 1. The first bolometers Bol.sub.1(i,j) are arranged in the corners of the detector in this configuration. The invention is however not limited to this layout, and the first bolometers Bol.sub.1(i,j) may for example form complete columns of bolometers, in particular the columns bordering the matrix-array of bolometers.

[0065] It is also understood that a first bolometer Bol.sub.1(i,j), since it is masked, is sensitive only to the radiation associated with the diaphragm that closes it off during the acquisition of data, this diaphragm advantageously being at the same temperature as the detector 1. The detector in FIG. 2 may also comprise a plurality of blind bolometers Bol.sub.BLIND(j).

[0066] The detector 1 furthermore comprises a computer 4 provided with a computing processor intended to execute the various steps of the method according to the present invention. The computer may also comprise a memory space for saving the raw measurements, reference measurements or intermediate measurements.

[0067] Lastly, the detector 1 may comprise a temperature probe 5 intended to evaluate a reference temperature, in particular the temperature of the detector, denoted T.sub.amb. The temperature probe may for example comprise a PN junction.

[0068] The method according to the present invention proposes to correct the defects in a raw image collected by the detector 1. The raw image is characterized in particular by a set of raw measurements Pix(i,j) representative of the temperature T.sub.scene of the scene observed by each bolometer i, j. These raw measurements Pix(i,j) are derived from a signal S.sub.p(i,j) actually delivered by each bolometer and representative of the observed scene via the equation:

S.sub.P(i,j)=R.sub.esp(i,j)(T.sub.scene−T.sub.amb)+S.sub.0,T.sub.amb  [Math 4]

[0069] Where: [0070] T.sub.amb is the temperature of the imager; [0071] T.sub.scene is the temperature of the scene seen by the bolometer i, j; [0072] R.sub.esp(i,j) is the responsiveness of the bolometer; [0073] S.sub.0,Tamb is the output value of the bolometer for a scene temperature equal to the ambient temperature.

[0074] It is understood that the temperature T.sub.amb is the temperature of each of the bolometers of the imager, and that this is the same as the ambient temperature in the case of thermal equilibrium with the surroundings in which the imager is located.

[0075] The responsiveness R.sub.esp(i,j) of each of the bolometers Bol(i,j) may be determined during production of the detector or before it is first used. In particular, determining the responsiveness Bol(i,j) may involve a first and a second calibration measurement. The first calibration measurement may in particular comprise exposing the detector, and more particularly all of the bolometers, to an object or black body kept at a first temperature that is advantageously equal to that of said bolometers and is denoted T.sub.amb below. The second calibration measurement may comprise exposing the detector a second time to another black body kept at a second temperature that is different from, for example higher than, the first temperature. Simply knowing the reference temperature, in particular the temperature of the detector, as well as the first and second temperatures makes it possible to deduce the responsiveness R.sub.esp(i,j) of each of the bolometers Bol(i,j). The responsivenesses R.sub.esp(i,j) are in particular saved in the memory space of the computer 4.

[0076] This calibration procedure remains easy to execute, and does not require implementing algorithms that consume a large amount of memory space and significant computing resources.

[0077] FIG. 1 shows one example of an image formed from raw measurements Pix(i,j). This FIG. 1, discussed in the “PRIOR ART” section, exhibits uniformity defects and a column-like appearance.

[0078] The non-uniformity is essentially caused by a dispersion in the electrical resistances of the bolometers forming the detector.

[0079] Darker regions may also be observed in the corners of the image. These correspond to the raw measurements Pix(i,j), called first raw measurements Pix.sub.1(i,j), collected by the first bolometers Bol.sub.1(i,j). These dark regions result from the first bolometers Bol.sub.1(i,j) being closed off by the diaphragm on which the lens is mounted, and will be utilized to correct the defects actually observed in the raw image.

[0080] A brighter central region, corresponding to the raw measurements Pix(i,j), called second raw measurements Pix.sub.2(i,j), of the second bolometers Bol.sub.2(i,j) may also be seen.

[0081] Finally, FIG. 1 also exhibits a column-like appearance, which is caused by the dispersion in the electrical resistances of the blind bolometers Bol.sub.BLIND(j). In this respect, each blind bolometer Bol.sub.BLIND(j) is implemented for the differential measurement of the bolometers of at least one column of bolometers that is specific thereto; each blind bolometer Bol.sub.BLIND(j) is advantageously associated with a single column C.sub.j of bolometers Bol(i,j).

[0082] It should be noted that, if all of the columns were to be associated with the same blind bolometer, the column effect would not be observed. However, this configuration is undesirable since any defect with this single blind bolometer would render the whole detector inoperative.

[0083] The method according to the present invention comprises a correlation step a) between the first raw measurements Pix.sub.1(i,j) and first reference measurements Pix.sub.1REF(i,j). The first reference measurements Pix.sub.1REF(i,j) are in particular measurements collected by each of the first bolometers Bol.sub.1(i,j) by applying a mask or a shutter, covering the whole of the detector 1 (and therefore all of the first and second bolometers), and kept at a reference temperature, for example the same temperature as said detector 1, such as 20° C.

[0084] The mask or the shutter should not be confused with the abovementioned diaphragm. Specifically, the mask is a part that is placed in front of the detector (the bolometers) for calibration purposes or in order to determine technical features of the bolometers during production thereof. No consideration whatsoever is given to this mask when using the detector.

[0085] It is understood that, when acquiring first reference measurements Pix.sub.1REF(i,j), second reference measurements Pix.sub.2REF(i,j) are also acquired by the second bolometers Pix.sub.2REF(i,j). These second reference measurements Pix.sub.2REF(i,j), like the first reference measurements Pix.sub.1REF(i,j), correspond to the image of a scene kept at a reference temperature, for example at the same temperature as the detector. These second reference measurements Pix.sub.2REF(i,j) are in particular used during step b) according to the invention described above.

[0086] The first reference measurements Pix.sub.1REF(i,j) and the second reference measurements Pix.sub.2REF(i,j) may be saved in the memory space of the computer 4. These reference measurements may be acquired in the course of a calibration procedure and advantageously be acquired at the same time as the first calibration measurement described above.

[0087] Correlation step a) according to the present invention may involve a linear regression. In particular, correlation step a) may involve determining coefficients α and Pix.sub.offset through linear regression, in particular based on first reference measurements Pix.sub.1REF(i,j), such that:

Pix.sub.1(i,j)=α.Math.Pix.sub.1REF(i,j)+Pix.sub.offset  [Math 5]

[0088] FIG. 3 shows a graph of the first raw measurements Pix.sub.1(i,j) (vertical axis) as a function of the first reference measurements Pix.sub.1REF(i,j) (horizontal axis). A linear trend is clearly observed, thus making it possible to justify the choice of a correlation by way of a linear regression.

[0089] This aspect is unexpected since no technical element makes it possible to foresee such behaviour. On the contrary, a more erratic trend was expected, given the plurality of technical parameters defining the operation of a bolometer.

[0090] The correlation through linear regression is obtained in particular without any prior correction or calibration of the column-like appearance of the first reference measurements Pix.sub.1(i,j), thereby still making it possible to obtain a satisfactory processing method that thus requires fewer computing resources.

[0091] The method according to the present invention also comprises a step b) of correcting raw measurements Pix(i,j).

[0092] Step b) of correcting the raw image comprises in particular computing corrected measurements Pix.sub.Cor(i,j) of a corrected image for each bolometer Bol(i,j) on the basis of the reference measurements Pix.sub.REF(i,j) and of the result of correlation step a).

[0093] In particular, step b) may involve computing the corrected measurements Pix.sub.Cor(i,j). This correction of the raw measurements may involve all of the raw measurements Pix(i,j) or only the second raw measurements Pix.sub.2(i,j) representative of the scene. It should be noted that this second alternative remains more advantageous in terms of computing resources.

[0094] The corrected measurements Pix.sub.Cor(i,j) may be computed by way of the following relationship:

Pix.sub.Cor(i,j)=Pix(i,j)−α.Math.Pix.sub.REF(i,j)−Pix.sub.offset  [Math 6]

[0095] Computing the corrected measurements Pix.sub.Cor(i,j) therefore depends on the reference measurements Pix.sub.REF(i,j), without requiring any prior correction or calibration thereof. In this respect, FIGS. 4a and 4b illustrate the effect of the correction according to the present invention on an image collected from a scene by the detector 1.

[0096] In particular, FIG. 4a is a raw image of a scene, representing an individual, collected by the detector 1. The pixelation and column effects however do not allow the individual to be distinguished in this image.

[0097] FIG. 4b, for its part, shows an image corrected in accordance with the terms of the present invention. This image, which no longer exhibits pixelation and a column-like appearance, allows the individual to be distinguished very clearly.

[0098] The method according to the present invention thus makes it possible to correct the defects inherent to images formed by way of bolometer detectors. The method according to the present invention makes it possible in particular to correct the pixelation and column effects commonly observed in images formed by way of bolometer detectors.

[0099] The method according to the present invention, which exhibits relatively low consumption in terms of computing resources, remains highly robust and efficient. Moreover, this method does not involve any additional mechanical part provided with motorization means, such that its manufacturing and usage cost is thereby reduced. The invention also relates to a computer program that, when it is executed by a computer, leads to the correction method according to the present invention being implemented.

[0100] The invention also relates to a device comprising: [0101] a detector provided with bolometers Bol(i,j) arranged in the form of a matrix-array of n rows (L.sub.i) and m columns (C.sub.i), the bolometers comprising first Bol.sub.1(i,j) and second Bol.sub.2(i,j) bolometers, the first bolometers Bol.sub.1(i,j) being closed off; [0102] a computer provided with the computer program according to the present invention.

REFERENCES

[0103] [1] EP294099161. [0104] [2] US2010237245A1. [0105] [3] EP 3594643A1.