Method for coloring infrared images

Abstract

The method for coloring an infrared image (11) comprises the following steps: - a step of acquiring an infrared image (11) obtained from an infrared sensor (10) comprising a set of elementary detectors, said infrared image (11) comprising, for each elementary detector, pixels having a value coded in a gray scale (NG) corresponding to a value representative of the infrared radiation received by said elementary detector; generating a colored image (13) by relating each pixel value coded in grayscale (NG) to a pixel value coded in color space by means of a color palette (12); Including, The color palette (12) is defined in CIE CAM02-UCS space by a palette generation module (20) using a parameter orbit, the parameter being the polar angle vector (Φ) of the parameter orbit, Hue and chromaticity plane (O, a * , b * ) the parameter orbits projected onto the Perceived luminance (j) in relation to grayscale value (NG) * ) dynamic range follows one or two straight lines It is stipulated as follows.

Description

[Technical Field]

[0001] The present invention relates to the field of electromagnetic radiation detection, and more particularly to infrared radiation detection.

[0002] The invention relates more particularly to the problem of displaying infrared images, i.e. images resulting from the detection of infrared radiation, and more particularly to coloring infrared images in order to improve their interpretation by the human brain.

[0003] Thus, the present invention can be implemented in many fields where thermal imaging is currently used, such as aerospace, security, defense, transportation, thermography, industrial inspection, building inspection, leisure activities, health, etc. [Background technology]

[0004] In the health field, infrared detectors have been widely developed since the COVID-19 pandemic. Indeed, people suffering from this disease are traditionally found to have elevated temperatures, and infrared detectors make it possible to obtain images illustrating the temperatures of objects present in the scene. Thus, infrared detectors are specifically used to identify the temperature of people passing through quarantine areas in order to detect whether a person has a high fever and is therefore at risk of being COVID-19 positive.

[0005] Infrared detectors are conventionally capable of obtaining images that are reconstructed in grey scale, where each pixel represents the amount of radiation captured by the underlying detector that forms a pixel of the infrared image.

[0006] However, this grayscale representation makes it difficult to visually distinguish the temperature difference between two objects in the scene that have similar temperatures.

[0007] For example, in the case of detecting people with COVID 19, this grayscale representation is often insufficient to allow an operator to quickly detect people with high temperatures in a group of people.

[0008] It is therefore desirable to improve this representation by coloring the representation of the amount of radiation captured by the underlying detectors that form the pixels of an infrared image.

[0009] This research is particularly motivated by the fact that studies have shown that in trained Westerners, the human eye can only perceive a limited number of shades of gray, between 120 and 180 shades, while humans can perceive up to 2 million colors. Therefore, by appropriately selecting colors, it is possible to significantly improve the perception of detail in infrared images.

[0010] For this purpose, it is known from documents WO 2016 / 179050, US Pat. No. 10,298,859 or WO 2014 / 200586 to convert a grey-level infrared image into a coloured image by applying a colour palette, which makes it possible to associate for each greyscale coded pixel value a pixel value coded in colour space.

[0011] To define the color space, the light spectrum W(λ) of a scene is considered in units of power, where λ is the wavelength in micrometers. In the case of a real scene in visible light, W(λ) is the fraction reflected, diffused, or transmitted from a scene illuminated by an external light source of a given spectral length, preferably a white light source, i.e., a broad band in the visible range. In the case of a display by a screen, W(λ) is the additive linear combination of the emission spectra of the triplet of elementary light sources: red, green, and blue pixels.

[0012] The human eye perceives brightness and color using rods and cones in the retina. In low light conditions, typically scotopic vision, rods provide a unique spectral response. In high light conditions, typically photopic vision, cones are responsible for color perception and are classified according to three groups: short, intermediate, and long, which have distinct spectral responses centered at short, intermediate, and long wavelengths.

[0013] Thus, to the human eye, color is characterized by simple scalar triplets, not the full scene spectrum. This argument makes it possible to explain how color can be properly rendered through a screen formed from three fundamental light sources of red, green, and blue spectral length.

[0014] In the sRGB space for "standard red-green-blue", colors are represented by red, green, and blue triplet colors, typically coded over a depth of three times 8 bits, or 24 bits.

[0015] Over time, other spaces have been defined to achieve more accurate color representations. For example, CIE XYZ is a coordinate system for color space defined by the International Commission on Illumination (CIE) in 1931. CIE XYZ is derived from the sRGB space and assumes that the relationship between physical quantities and perception is linear. This approximation places sets of points representing colors in a three-dimensional vector space. To define this CIE XYZ space, color experiments were required, where people were asked to adjust the ratios of three experimental primaries to obtain the same color point as the one being evaluated. The sum of the "X," "Y," and "Z" components represents absolute luminance, but not the actual perceived luminance.

[0016] The International Commission on Illumination (CIE) also defined the CIE LAB space, based on the CIE XYZ space, in 1976. The CIE LAB space is a color space used specifically for characterizing surface colors.

[0017] Three quantities characterize a color: lightness L, derived from the luminance of a surface, and two parameters a and b, describing the color's difference from that of a gray surface of the same lightness. The existence of a gray, colorless, achromatic surface implies that the composition of the light illuminating the colored surface is explicitly stated. This "illuminant" is often normalized daylight.

[0018] The CIE LAB space has the major drawback of inaccurately modeling human perception of luminance and hue. In other words, displacements by the basic Euclidean lengths in the CIE LAB space do not necessarily result in the same color perception distances, especially in terms of luminance and hue. This can be explained by the fact that the model does not take into account principles of perceptual chromatic adaptation, such as local luminance and hue adaptation.

[0019] More recently, in 2002, the International Commission on Illumination (CIE) defined the CIE CAM02 space for calculating the mathematical correlations that exist between six technical dimensions linked to color appearance: brightness, luminance, chroma, chromaticity, saturation, and hue.

[0020] This model is defined based on the X, Y, and Z quantities of the CIE XYZ space by considering the visualization of a reference stimulus. This stimulus corresponds to a reference white point. The CIE CAM02 model also takes into account the background against which the stimulus is viewed, information about the luminance conditions of the stimulus's environment, and information about whether or not the luminance conditions are taken into account by the observer of the stimulus. This model makes it possible to predict the color appearance of a stimulus with these attributes or to calculate the corresponding colors under different viewing conditions.

[0021] Other color spaces also exist, such as the HSV space for "hue-saturation value" or the HSL space for "hue-saturation-luminance".

[0022] For example, document WO2014 / 200586 provides a method for generating a color palette based on interpolation in HSL space. Based on a set of predefined hue values, the palette is constructed with constant brightness, fixed saturation, and linear interpolation to hue. Depending on the type of scene, this method makes it possible to use an adapted range of hues, for example a red-orange-yellow palette for a desert scene or a yellow-green-blue palette for a forest scene.

[0023] In document WO2016 / 179050, a color palette is generated by interpolation in sRGB space based on a set of base colors corresponding to interpolation nodes.

[0024] Document U.S. Pat. No. 10,298,859 discloses a color solution in sRGB, CIE XYZ, or CIE LAB space, primarily by using a predefined color palette for a specific application. In one embodiment, the color palette is generated based on the minimum and maximum grayscale values ​​of the infrared image. This color palette is thus adapted to the dynamic range of the observed scene.

[0025] Thus, scenes with low thermal contrast are shown with a low hue dynamic range that is compensated for by an adapted luminance dynamic range. This document also provides for improving the perceived contrast by using a luminance filter that exhibits oscillations.

[0026] Whatever the method used, the goal is to obtain a color palette on a monochromatic image by adhering to the color sequence represented by the sRGB triplet while adhering to certain rules prescribed by a specification. Such specifications traditionally include both cognitive and communicative criteria, which may be contradictory.

[0027] The criteria for communication relate to aesthetic appearance, compliance with special permits on the drawing, or even the search for conformity to a particular evocative use, for example the desired colors to represent animals in a hunting image.

[0028] Cognitive criteria aim, for example, to facilitate the interpretation of an image according to its scene type, to improve the perception of contrast in an image, to enhance the characteristics of the object type of the subject, or to facilitate the interpretation of the image for people with visual impairments, typically people who are color blind.

[0029] In all cases, the threshold for recognition varies from person to person.

[0030] However, existing solutions are capable of designing color palettes that are linear in absolute luminance, but are not linear in terms of perception with respect to a particular person's cognitive criteria. As a result, certain details in a colorized infrared image are not necessarily visible to a given person, while they are visible to another person. [Prior art documents] [Patent documents]

[0031] [Patent Document 1] WO2016 / 179050 [Patent Document 2] U.S. Patent No. 10,298,859 [Patent Document 3] WO2014 / 200586 [Non-patent literature]

[0032] [Non-Patent Document 1] R. Luo, G. Cui, and C. Li, "Uniform color spaces based on CIECAM02 color appearance model," Color Research and Application, 2006 Summary of the Invention

[0033] Therefore, the technical problem of the present invention is to obtain a method for coloring an infrared image with a linear color palette according to human perception, so that a person can easily change the color palette for a specific application and / or for personal perception. [Means for solving the problem]

[0034] To address this technical problem, the present invention provides a color palette defined in the CIE CAM02-UCS space by a parameter orbit. This uniform CIE CAM02-UCS color space is defined in the article by R. Luo, G. Cui, and C. Li, "Uniform color spaces based on CIECAM02 color appearance model," Color Research and Application, 2006. The axis j of this Cartesian space is * , a * , and b * So, j * corresponds to the perceived luminance, and a * corresponds to the blue-yellow color axis, and b * corresponds to the green-red color axis.

[0035] Chromaticity is defined by the following formula:

[0036]

number

[0037] Hue is defined by the following formula:

[0038]

number

[0039] More specifically, the invention arises from the observation that, to obtain effective linearity in terms of perception, the parameters of this trajectory must be polar angle vectors. Following this same observation, the trajectory projected onto the hue and chromaticity plane must also follow one or two ellipses, while the dynamic range of perceived luminance with respect to grayscale values ​​must follow one or two straight lines.

[0040] In the present invention, the user can use predefined mathematical definitions of this parameter trajectory for different applications to obtain a color palette that is perceptually linear with respect to common cognitive standards.

[0041] Furthermore, the user may also adjust certain parameters of this parameter trajectory to adapt the general cognitive criteria to his / her own sense.

[0042] The present invention therefore relates to a method for coloring an infrared image, comprising the following steps: - acquisition of an infrared image obtained from an infrared sensor comprising a set of elementary detectors, said infrared image comprising, for each elementary detector, pixels having a greyscale coded value corresponding to a value representative of the infrared radiation received by said elementary detector; - generating a colored image by associating each grayscale coded pixel value with a color space coded pixel value by means of a color palette; The present invention relates to a method for coloring an infrared image, including:

[0043] The present invention provides a module for generating a palette using a parameter orbital whose parameters are formed by the polar angle vectors of the parameter orbital, the color palette being defined in the CIE CAM02-UCS space, and the parameters being: - The parameter trajectories projected onto the hue and chromaticity plane follow one or two ellipses, - The dynamic range of perceived luminance with respect to grayscale values ​​follows one or two straight lines The present invention is characterized in that it is defined as follows.

[0044] Preferably, the parameter trajectory obeys the following relationship:

[0045]

number

[0046] Here, the first semi-axis (a ell ) divided by 2 is the peak chromaticity (c c * ) and the second semi-axis (b ell ) is calculated by the following equation for the first semi-axis (a ell ) is defined as a function of

[0047] If the eccentricity (e) of the ellipse is positive, then

[0048]

number

[0049] If the eccentricity (e) of the ellipse is negative, then

[0050]

number

[0051] With these precisely constructed parameters, the present invention is able to obtain a color palette linearly designed in the CIE CAM02-UCS space, which ensures good luminance perception linearity.

[0052] The color palette thus defined can then be modified by the user according to his / her own expectations. For this purpose, the method comprises the following steps: - displaying said colored image on a screen viewable by a user; - From the man-machine interface, the recovery steps of the color changes expected by the user and It can also include:

[0053] Preferably, the human-machine interface is capable of varying the peak hue, peak chromaticity, the dynamic range of the polar angle vector, the eccentricity of the at least one ellipse projected onto the hue and chromaticity plane, and / or the dynamic range of the perceived luminance, which parameters allow the color palette to be altered in ways that will be readily apparent to those skilled in the art who wish to improve the chromaticity of an image.

[0054] In the sense of the present invention, the hue or chromaticity peak refers to the point at which the polar angle corresponds to π. If the ellipticity is positive, this peak corresponds to the chromaticity maximum.

[0055] Preferably, the dynamic range of the polar angle vector is limited to between 0 and 2π or 2π and 0.

[0056] Additionally, to simplify the selection of a color palette for a particular application, the human-machine interface may allow for the selection of a predefined palette.

[0057] For example, to evoke the idea of ​​thermal imaging, while avoiding unaesthetic purple hues and overly intense white and yellow hues, the first predefined palette preferably has a peak hue in the range of 20 to 30°, a peak chromaticity in the range of 30 to 40, a dynamic range of the polar angle vector in the range of 0 to 1.8π, an eccentricity of the ellipse projected onto the hue and chromaticity plane in the range of −0.35 to −0.25, and a dynamic range of perceived luminance in the range of 5 to 100. To represent warm colors, this first predefined palette avoids pure white and overly bright yellow, limiting aggressiveness and visual fatigue effects.

[0058] In night vision applications, the green monochromatic display of nighttime images is based on phosphorus scintillators used in photoluminescence tubes, which have a spectral re-emission length centered in green. Because green is particularly conveyed by military propaganda via motion pictures and video games, the use of green has become a public symbol and association for nighttime images. The use of green is also advantageous in terms of light sensitivity, since the spectral sensitivity curve, at least in good lighting conditions, has a maximum around 555 nanometers.

[0059] In this type of night vision application, the display is generally set at a very low brightness, which on the one hand makes it possible to reduce the time for the user's eyes to adapt when they stop looking at the screen displaying the tinted infrared image to directly observe their environment, and on the other hand makes it possible to reduce the light reflected on the face for stealth reasons.

[0060] Now, in scotopic vision, the spectral sensitivity curve is adapted to darkness by including rods rather than cones, and so in this case scotopic vision has a sensitivity curve centered around 507 nanometers.

[0061] The second predefined palette is designed to facilitate viewing infrared images in low-light conditions while limiting visual fatigue when viewing video streams for extended periods of time. To this end, the second predefined palette has a peak hue ranging from -170 to -160°, a peak chromaticity ranging from 10 to 20, a dynamic range of the polar angle vector ranging from 0 to 2π, an eccentricity of the ellipse projected onto the hue and chromaticity plane ranging from 0.2 to 0.4, and a dynamic range of perceived luminance ranging from 100 to 20.

[0062] With this second predefined palette, the image appears smooth and rounded, while remaining sharp and prominent on the target objects. The background, especially the sky, appears in light colors, which reduces both pupil aperture and interpretation effort due to familiarity with daytime scenes, especially when the sky is clear. The aesthetic appearance of the colorized image provides the appearance of a glossy print, which is considered rather pleasant.

[0063] The third predefined palette is formed by a threshold palette, i.e., a palette constructed by concatenation of two palettes, with a grayscale threshold at which the transition between the two palettes occurs. Preferably, color continuity is ensured at the level of said threshold. These two palettes allow the separation of two groups / classes of pixels by coloring them differently and maximizing the dynamic range of perceived luminance of each class. These two groups / classes of pixels may differ by distinct hue or chromaticity variations.

[0064] To distinguish the hot areas from the ambient temperature of the scene by maximizing the detail of the two areas, the third predefined palette preferably includes a first portion having a peak hue in the range of -110 to -100°, a peak chromaticity in the range of 15 to 20, a dynamic range of the polar angle vector in the range of 2π to 0, an eccentricity of the first ellipse projected onto the hue and chromaticity plane in the range of 0.5 to 0.7, and a dynamic range of the perceived luminance in the range of 100 to 5, and then a second portion having a peak hue in the range of 25 to 40°, a peak chromaticity in the range of 30 to 40, a dynamic range of the polar angle vector in the range of 0 to 1.8π, an eccentricity of the second ellipse projected onto the hue and chromaticity plane in the range of 0.3 to 0.4, and a dynamic range of the perceived luminance in the range of 5 to 99, with the threshold between the two portions being set between 45% and 55% of the grayscale value.

[0065] In this third predefined palette, the representation of the target object is achieved by a palette close to the first predefined palette, but with the difference that the dynamic range of hues is slightly reduced to avoid overloading the image.

[0066] Thus, the objects of interest are colored normally, which facilitates interpretation of the image properties. The perceived brightness has a sufficiently increased dynamic range, ensuring linearity of color changes for the user.

[0067] The background is also represented via an elliptical palette, with a significantly reduced dynamic range and a slate-blue peak in perceived brightness. This peak color is selected at a low chromaticity because it is desirable to draw attention to the target object and not the background. Similarly, in this third predefined palette, it is desirable to limit hue variation via a more elongated elliptical orbit. These choices also result in reduced sensitivity to small intensity variations, partially limiting the perception of noise; the signal-to-noise ratio is generally lower on the surrounding background than on the target object.

[0068] The background can be brightened by reducing its perceived luminance to improve the quality of the target object. This quality is emphasized by a black outline at the threshold and by the chromaticity contrast between the two classes of pixels. For this purpose, pixels around the threshold are typically colored with a dark color with low chromaticity, such as a color close to black. These pixels are generally at the boundary between the background and the target object.

[0069] Finally, the use of blue to represent the background can be justified on the one hand because it is a hue associated with cool temperatures, and on the other hand because it is a hue associated with the sky, which is a comforting representation for humans. For the same reason, the use of blue creates the illusion of depth by recalling the horizon, which is at least the technique used by impressionists to create this effect.

[0070] In a fourth predefined palette, it is desirable to define a palette that includes grayscale portions across the entire dynamic range of perceived luminance and brightly colored portions with high-threshold to red hues.

[0071] In hunting applications, hot spots are typically animals, and it is desirable to highlight these in the image. In such applications, the environment is also very important. The environment typically consists of complex elements such as vegetation, making it important to represent the environment in its full grayscale dynamic range. Other applications may need to alert the user, for example, ADAS, an "Advanced Driver Assistance Systems" application that consists of alerting the driver to the presence of pedestrians for braking assistance.

[0072] For this purpose, the fourth predefined palette preferably has a first portion that is free of chromaticity and has a dynamic range of perceived luminance ranging from 0 to 100, and then a second portion that has a peak hue ranging from 30 to 35°, a peak chromaticity ranging from 30 to 40, a dynamic range of the polar angle vector ranging from 2π to 0, an eccentricity of the ellipse projected onto the hue and chromaticity plane ranging from 0.8 to 0.9, and a dynamic range of perceived luminance ranging from 100 to 30, with the threshold between the two portions being set between 75% and 85% of the grayscale value.

[0073] The manner in which the invention can be implemented, as well as the resulting advantages, will become apparent from the following embodiments given by way of non-limiting indication, in connection with FIGS. [Brief explanation of the drawings]

[0074] [Figure 1] 1 is a schematic illustration of a method for coloring an infrared image according to an embodiment of the present invention. [Figure 2] 2 is a schematic diagram of the parameter generation of a color palette according to the method of FIG. 1, with an ellipse projected onto the hue and chromaticity plane. [Figure 3] 2 shows three variants of parameter generation of a color palette according to the method of FIG. 1, with an ellipse projected onto the hue and chromaticity plane according to the variation of the eccentricity of the ellipse. [Figure 4] FIG. 2 shows the perceived luminance and dynamic range of the ellipse projected onto the hue and chromaticity plane for the first predefined palette of the method of FIG. 1. [Figure 5] 10A and 10B show the perceived luminance and dynamic range of the ellipse projected onto the hue and chromaticity plane for the second predefined palette of the method of FIG. 1. [Figure 6] 10A and 10B show the dynamic range of perceived luminance and the two ellipses projected onto the hue and chromaticity planes for the third predefined palette of the method of FIG. 1. [Figure 7] FIG. 2 shows the dynamic range of perceived luminance and the projected ellipses onto the hue and chromaticity planes for the fourth predefined palette of the method of FIG. 1. DETAILED DESCRIPTION OF THE INVENTION

[0075] Figure 1 illustrates a method for colouring an infrared image 11 obtained from an infrared sensor 10 comprising a set of elementary detectors. Conventionally, the elementary detectors are formed from microbolometers arranged in an array.

[0076] For each row or column of microbolometers, a readout circuit makes it possible to measure the amount of infrared radiation captured by each elementary detector, thus forming a pixel of the infrared image.

[0077] The value resulting from each pixel of the infrared image can be coded in grayscale NG over 8, 11, or 14 bits without altering the present invention. Furthermore, pre-processing can be performed on the infrared image 11 before implementing the coloring method of the present invention.

[0078] More specifically, the invention realizes the conversion, using a color palette 12, of an infrared image 11 coded in grayscale NG into a colorized image 13, each pixel of which is coded by a color system according to, for example, the sRGB standard.

[0079] In accordance with the present invention, the color palette 12 defines each color according to six characteristics: brightness, absolute luminance, chrominance, and chromaticity. * , saturation, and hue h * These six characteristics are interdependent and are defined in the CIE CAM02-UCS space by specifying * , a * , and b * and the parameter j * corresponds to the perceived luminance, and a * corresponds to the blue-yellow axis, and b * corresponds to the green-red axis.

[0080] In this CIE CAM02-UCS space, chromaticity c * is defined by the following formula:

[0081]

number

[0082] hue h * is defined by the following formula:

[0083]

number

[0084] More specifically, the present invention realizes the generation of a color palette 12 in the CIE CAM02-UCS space by a palette generation module 20 that uses a parametric trajectory whose parameter is the polar angle vector Φ. c * , b c * ) follows one or two ellipses, and the perceived brightness j with respect to the grayscale value NG * The definition of this parameter is constrained so that the dynamic range of follows one or two straight lines.

[0085] FIG. 2 shows the hue and chromaticity planes (O, a) of the parameter generation of the color palette 12. c * , b c * ) is illustrated. In this example, the ellipse is formulated according to a parametric orbit, whose parameters are length 2 n The polar angle vector φ = [0,...,φmax], where n corresponds to the depth of the digitized pixel value. Therefore, the various grayscale values ​​NG range from 0 to 2 n-1 Preferably, the depth φmax=1.8π. Preferably, the dynamic range of the polar angle vector Φ is limited to between 0 and 2π or 2π and 0.

[0086] In the example of Figure 2, the ellipse does not complete a full rotation, i.e., it is not closed, to avoid ending the palette with pure white, which would be considered too intense.

[0087] The ellipse is defined according to the following formula:

[0088]

number

[0089] where the ellipse is the peak chromaticity c divided by 2 in a reference frame aligned with the axes of the ellipse (O, a', b'). c *The first semi-axis a corresponds to ell and the second semi-axis b ell is the first semi-axis a ell and depends on the eccentricity e of the ellipse.

[0090] More specifically, the second semi-axis b ell can be estimated from the desired eccentricity e, if the eccentricity is positive, by the following equation:

[0091]

number

[0092] When the eccentricity is negative, the second semi-axis obeys the formula:

[0093]

number

[0094] When the eccentricity is positive, the first semi-axis a ell is the second semi-axis b ell While larger, when the eccentricity is negative, the first semi-axis a ell is the second semi-axis b ell Smaller than.

[0095] The effect of varying the eccentricity e is illustrated in Figure 3, where parameter trajectories with small eccentricities allow for the exploitation of purple and orange hues, while parameter trajectories with high eccentricities avoid such hues at the risk of making the image appear monochromatic.

[0096] For good visual appearance and to facilitate detailed perception through a good dynamic range of the purple / orange hues, it is preferable to choose a small eccentricity, typically an eccentricity in the range of -0.2 to -0.4, e.g., e=-0.3.

[0097] Whatever the value of eccentricity chosen, the angle h, according to the formula c *A simple azimuthal rotation of the reference frame (O, a) in Figure 2 by the hue of the desired peak. * , b * ) it is possible to perform a change of reference frame to draw an ellipse in

[0098]

number

[0099] The peak hue angle is h c * =atan2(b c * ,a c * ) and the perceived luminance curve j * (φ) is simply a line, and we can set the limit of its dynamic range. Peak hue h c * , peak chromaticity c c * , the dynamic range of the polar angle vector Φ, the eccentricity e, and / or the perceived luminance j * The value of the dynamic range of can be selected according to the type of color palette desired. For example, it is possible to increase the dynamic range of hues by decreasing the eccentricity e of the ellipse.

[0100] With these exact parameters of the definition of the color palette 12, it is possible to obtain a table that associates a triplet in sRGB space with each grayscale value N. For example, this conversion between the CIE CAM02-UCS space and the sRGB space can be performed by a "colorspacious" module coded in the Python computer language. This table is sometimes called a LUT for "look-up table."

[0101] In this color palette 12, each grayscale value NG is thus associated with a color triplet in sRGB space, so that the infrared image 11 can be converted into a colored image 13. The colored image 13 can now be displayed on a screen 14 so that a user can efficiently detect salient elements of the scene captured by the infrared sensor 10.

[0102] To detect specific elements for various applications, the user can use the man-machine interface 15 to select a predefined color palette.

[0103] Additionally, the man-machine interface 15 may also allow the user to adjust the configuration values ​​of the color palette 12 .

[0104] In fact, color palette 12 is the peak hue h c * , peak chromaticity c c * , the dynamic range of the polar angle vector Φ, the eccentricity e, and / or the perceived luminance j * Even if the dynamic range is specified by values ​​of 100,000, the user can optionally change these values ​​via the man-machine interface 15. To this end, the user can request an increase or decrease of each value in order to observe the progress of these changes directly on the screen 14 for viewing the colored image 13.

[0105] With regard to predefined color palettes, four palettes may be offered to the user via the man-machine interface 15: Lifeinred™ Color, Lifeinred™ Serenity, Lifeinred™ Contrast, and Lifeinred™ Tracker.

[0106] For example, the predefined palette Lifeinred™ Color has peak hues in the range of 20-30°. c *, peak chromaticity c in the range of 30-40 c * , the dynamic range of the polar angle vector Φ ranging from 0 to 1.8π, the hue and chromaticity planes ranging from -0.35 to -0.25 (0, a * , b * ) and the eccentricity e of the ellipse projected onto the * It has a dynamic range of

[0107] As illustrated in FIG. 4, this predefined palette Lifeinred™ Color provides a range of perceived luminance j * and the linearly increasing dynamic range of the hue and chromaticity planes (O, a * , b * ) contains an ellipse within the rectangle. Point P1 corresponds to the color coral red, point P2 corresponds to the color vanilla or peach, and point P3 corresponds to the color wine red.

[0108] The predefined palette Lifeinred™ Serenity has peak hues ranging from -170 to -160°. c * , peak chromaticity c in the range of 10 to 20 c * , the dynamic range of the polar angle vector Φ in the range of 0 to 2π, the hue and chromaticity planes in the range of 0.2 to 0.4 (0, a * , b * ) and the eccentricity e of the ellipse projected onto the sphere, and the perceived brightness j ranging from 100 to 20. * It has a dynamic range of

[0109] As illustrated in FIG. 5, this predefined palette Lifeinred™ Serenity provides a perceived luminance j with a reduced dynamic range Dr. * The straight and reduced dynamic range of the hue and chromaticity planes (O, a * , b * ) Point P6 corresponds to the color water green, while point P5 corresponds to the color slate gray, and point P7 corresponds to the color teal blue.

[0110] The predefined palette Lifeinred™ Contrast provides peak hues in the range of -110 to -100°. c * , peak chromaticity c in the range of 15-20 c * , the dynamic range of the polar angle vector Φ is in the range of 2π to 0, and the plane (0, a * , b * ) and the perceived luminance j in the range of 100 to 5. * The first portion has a dynamic range of

[0111] The predefined palette Lifeinred™ Contrast provides peak hues in the 25-40° range. c * , peak chromaticity c in the range of 30-40 c * , the dynamic range of the polar angle vector Φ is in the range of 0 to 1.8π, and the plane (0, a * , b * ) and the eccentricity e of the second ellipse projected onto the * A threshold S1 between the two parts is set between 45% and 55% of the greyscale value NG, for example 50%.

[0112] As illustrated in FIG. 6, this predefined palette Lifeinred™ Contrast provides a perceptual luminance j decreasing to a threshold S1. * The straight dynamic range of j and then the perceived luminance j * The linear and increasing dynamic range of the two ellipses is formed in the plane (0, a*, b*). Point P1 corresponds to the color Coral Red, also used in the predefined palette Lifeinred™ Color, while point P5 corresponds to the color Slate Blue.

[0113] The predefined palette Lifeinred™ Tracker has a first part that is chromaticity-free and has a dynamic range of perceived luminance j* ranging from 0 to 100, and then a peak hue h ranging from 30 to 35°. c * , peak chromaticity c in the range of 30-40 c * , the dynamic range of the polar angle vector Φ is in the range of 2π to 0, and the plane (0, a * , b * ) and the eccentricity e of the ellipse projected onto the sphere, and the perceived brightness j ranging from 100 to 30. * and a second portion having a dynamic range of 1000 sigma. A threshold S2 between the two portions is set between 75% and 85% of the grayscale value, for example 80%.

[0114] As illustrated in FIG. 7, this predefined palette Lifeinred™ Tracker scales perceived luminance j up to a threshold S2. * The straight dynamic range of j and then the perceived luminance j * and a linearly decreasing dynamic range of 1000 to 12000. An ellipse is formed in the plane ab, where point P8 corresponds to brick red.

[0115] These various predefined color palettes and the man-machine interface 15 allow the user to change the parameters of the generation of the color palette 12, so the present invention allows the user to obtain a linear aspect according to his / her personal perception, thus enabling the user to more clearly observe the target object captured by the infrared detector. [Explanation of symbols]

[0116] 10 Infrared sensor 11. Infrared imaging 12-color palette 13 Colored images 14 screens 15 Man-machine interface 20 Palette Generation Module

Claims

1. A method for coloring an infrared image (11), comprising: - a step of acquiring an infrared image (11) obtained from an infrared sensor (10) comprising a set of elementary detectors, said infrared image (11) comprising, for each elementary detector, pixels having a value coded in a gray scale (NG) corresponding to a value representative of the infrared radiation received by said elementary detector; generating a colored image (13) by relating each pixel value coded in grayscale (NG) to a pixel value coded in color space by means of a color palette (12); In a method comprising: The color palette (12) is defined in CIE CAM02-UCS space by a module (20) for generating a palette using the parameter orbitals, the parameter being the polar angle vector (Φ) of the parameter orbitals, and the parameter being: Hue and chromaticity plane (O, a * , b * ) the parameter trajectory projected onto the The perceived luminance (j * ) dynamic range follows one or two straight lines A method for coloring an infrared image, characterized in that:

2. The parameter trajectory satisfies the following relationship: [Equation 1] in accordance with Here, the first semi-axis (a ell ) divided by 2 is the peak chromaticity (c c * ) and the second semi-axis (b ell ) is expressed as follows: If the eccentricity (e) of the ellipse is positive, then: [Equation 2] and if the eccentricity (e) of the ellipse is negative, then: [Equation 3] The first semi-axis (a ell 2. The method of claim 1, wherein the coloring of an infrared image is defined as a function of

3. displaying said colored image (13) on a screen (14) viewable by a user; a step of recovering the color changes expected by the user from the man-machine interface (15); 3. The method of coloring an infrared image according to claim 1 or 2, further comprising:

4. The man-machine interface (15) detects the peak hue (h c * ), peak chromaticity (c c * ), the dynamic range of the polar angle vector (Φ), the hue and chromaticity plane (0, a * , b * ) and / or the eccentricity (e) of at least one ellipse projected onto the * 4. The method for coloring an infrared image according to claim 3, wherein the method enables changing the dynamic range of the

5. 5. The method of coloring an infrared image according to claim 1, wherein the dynamic range of the polar angle vector (Φ) is limited between 0 and 2π or between 2π and 0.

6. 6. The method for coloring an infrared image according to any one of claims 3 to 5, wherein the man-machine interface (15) allows for the selection of a predefined palette.

7. The first predefined palette (Lifeinred™ Color) contains peak hues (h) ranging from 20 to 30°. c * ), peak chromaticity (c c * ), a dynamic range of the polar angle vector (Φ) ranging from 0 to 1.8π, and a range of the hue and chromaticity planes (0, a * , b * ) and the perceived brightness (j * 7. The method of claim 6, wherein the infrared image has a dynamic range of 100 s.p.m.

8. A second predefined palette (Lifeinred™ Serenity) provides peak hues (h) ranging from -170 to -160°. c * ), peak chromaticity (c c * ), the dynamic range of the polar angle vector (Φ) ranging from 0 to 2π, the hue and chromaticity planes (0, a * , b * ) and the perceived brightness (j * 8. The method of claim 6 or 7, wherein the method has a dynamic range of 100 sigma / s.

9. A third predefined palette (Lifeinred™ Contrast) provides peak hues (h) ranging from -110 to -100°. c * ), peak chromaticity (c c * ), the dynamic range of the polar angle vector (Φ) ranging from 2π to 0, the hue and chromaticity plane (0, a * , b * ) and the perceived brightness (j * ) dynamic range, and then a peak hue (h c * ), peak chromaticity (c c * ), a dynamic range of the polar angle vector (Φ) ranging from 0 to 1.8π, and a dynamic range of the hue and chromaticity planes (0, a * , b * ) and the eccentricity (e) of the second ellipse projected onto the image plane, and the perceived brightness (j) ranging from 5 to 100. * 9. The method of claim 6, wherein the first portion has a first dynamic range of 100% and the second portion has a second dynamic range of 100% and the threshold value (S1) between the first and second portions is set between 45% and 55% of the grayscale value (NG).

10. A fourth predefined palette (Lifeinred™ Tracker) is chromaticity-free and scales the perceived luminance (j) in the range 0 to 100. * ) dynamic range, and then a peak hue (h c * ), peak chromaticity (c c * ), the dynamic range of the polar angle vector (Φ) ranging from 2π to 0, the hue and chromaticity plane (0, a * , b * ) and the perceived brightness (j * 10. The method of claim 6, further comprising: a first portion having a first grayscale value (NG) and a second portion having a dynamic range of 100% to 200%; and a threshold value (S2) between the first and second portions being set between 75% and 85% of the grayscale value (NG).

Images