Identifying the visibility levels of visual information displayed on a screen

Abstract

Identification of Visibility Levels of Visual Information Displayed on a Screen. A method for determining a visibility indicator for visual information on a screen, when the visual information is displayed by the screen, characterized in that it comprises the following steps: (a) Obtaining an initial image of the screen when the screen displays the visual information, (b) Converting the initial image into a luminance image, (c) Detecting edges of the luminance image, the edges comprising a plurality of pixels, (d) Determining, for each pixel in the plurality of pixels, a visibility indicator of said pixel, by comparing the luminances of other pixels, the other pixels being determined from a position of said pixel in the luminance image. Figure for the abstract: Fig. 1

Description

Title of the invention: Identification of visibility levels of visual information displayed on a screen. Technical field

[0001] The present invention relates to the field of visual information display systems and more particularly to the evaluation and improvement of the visibility of digital, textual, or graphic visual information displayed on human-machine interface (HMI) screens. Previous technique

[0002] HMIs are commonly used in various sectors such as automotive, aeronautical, or electronic device design.

[0003] The visibility of visual information displayed on these HMIs is crucial for user safety and comfort, particularly in environments where lighting conditions can vary considerably. It is therefore necessary to be able to precisely determine the visibility levels of this visual information.

[0004] Prior solutions in this field include devices and methods for regulating the light intensity of indicator lights and displays.

[0005] Existing methods are often based on global contrast measurements or require manual interventions to assess visibility levels, which can be inaccurate and inefficient.

[0006] US patent 7639154 B2 describes a system for regulating the brightness of indicator lights according to the light intensity of the legends and displays, with separate controls for day and night modes.

[0007] We also know of patent KR100625755 Bl, which describes a form and character recognition device using color and luminance classification methods to obtain a quantification of high contrast and low contrast elements.

[0008] The technical problem to be solved is to automatically, spatially and precisely evaluate visibility indicators representative of a level of contrast of visual information displayed on a screen, in particular of a human-machine interface (HMI).

[0009] There is also a need to quickly identify areas of the screen that meet certain visibility criteria, particularly those related to existing standards, and to differentiate graphic elements from extraneous elements (reflection, sunlight, etc.).

[0010] Furthermore, existing solutions do not always take into account the user's perception of the screen. Nor do they allow for analysis of whether the contrast problem is related to the graphic content or to the usage conditions. Summary of the invention

[0011] The invention aims to meet all or part of these needs, and it achieves this, according to one of its aspects, by means of a method for determining a visibility indicator of visual information on a screen, when the visual information is displayed by the screen, and which comprises the following steps: a. Obtaining an initial image of the screen when the screen displays the visual information, b. Conversion of the initial image into a luminance image, c. Image edge detection in luminance, in other words, edge detection of image elements in luminance, the edges comprising a plurality of pixels, d. Determination, for each of the pixels of the plurality of pixels, of a visibility indicator of said pixel, by comparison of the luminances of other pixels, the other pixels being determined from a position of said pixel in the image in luminance.

[0012] Thanks to the invention, it is possible to obtain a spatial and objective representation of the visibility indicator at the pixel level, unlike existing methods which determine a global indicator, thus not taking into account local variations in the contrast levels of visual information.

[0013] Converting the initial image into luminance allows us to take into account the physiology of the human eye. Indeed, the three components red, blue, and green are not seen with the same intensity by the human eye. For example, the eye perceives green colors better than blue colors.

[0014] Edge detection on a luminance representation of the initial image, and not directly on the image or on a conversion of the image to grey levels, makes it possible to obtain a characterization of the edges relatively more faithful to the perception of the human eye.

[0015] In one embodiment, the method comprises determining one or more visibility indicators for all visual information on the screen. In other words, the method is implemented for all visual information on the screen. Display

[0016] The term "visual information of a screen" here refers to any visual information configured to be displayed by a screen. Visual information may include one or more textual, and / or numeric, and / or graphic elements.

[0017] In one embodiment, the visual information is a pictogram.

[0018] The method may include a step (e) of displaying an image representative of the visibility level indicator for each of the pixels of a selection of pixels from the plurality of pixels.

[0019] By "representative image of the visibility level indicator for each of the pixels of a selection of pixels", it is to be understood that the displayed image makes it possible to represent simultaneously, for each of the pixels of the selection of pixels, information representative of the visibility indicator of said pixel.

[0020] The display step makes it possible to obtain an objective, global, spatial and visual representation of the evaluated visibility indicator. Filtering

[0021] The pixels in the selection will be referred to hereafter as "selected pixels".

[0022] In one embodiment, the pixel selection is the set of pixels of the plurality of pixels, in particular the set of edge pixels.

[0023] The pixel selection may include only those pixels from the plurality of pixels whose visibility indicator is greater than a predetermined filtering threshold. Thus, it is possible to filter out "noise" in the displayed image or extraneous elements in the graphic elements, and / or to eliminate pixels whose visibility indicator does not reach a requirement threshold corresponding to the predetermined filtering threshold. Color levels

[0024] The method may include determining at least one visibility threshold, and for each of the selected pixels, pixels below the visibility threshold are not displayed, and pixels above the visibility threshold are displayed with a color code determined by different predefined representation models.

[0025] In one embodiment, the method offers different modes of representation such as contrast, contrast ratio, or visibility threshold. The first two modes of representation offer a linear representation based on a visibility threshold set by an operator of the invention. Below the set threshold, the contrast values ​​or contrast ratio can be represented by a gradient from a first color to a second color, and above the visibility threshold, the contrast values ​​or contrast ratio can be represented solely by the second color. This representation allows for the spatial visualization of elements that do not meet the operator's expectations.

[0026] In one embodiment, the method comprises determining a plurality of visibility thresholds defining disjoint ranges of values, the plurality of visibility thresholds comprising a minimum visibility threshold and a maximum visibility threshold, The value ranges covering all the values ​​of the visibility indicators of the selected pixels, the value ranges including: -a range of minimum values ​​comprising a set of values ​​below the minimum visibility threshold, -a range of maximum values ​​including a set of values ​​exceeding the maximum visibility threshold, and -at least one range of intermediate values ​​comprising a set of values ​​between two consecutive visibility thresholds,

[0027] the ranges of values ​​being associated with distinct respective colours,

[0028] the method comprising, for each of the selected pixels, the display of the selected pixel in the color associated with the range of values ​​including the visibility indicator of said pixel.

[0029] For example, the visibility indicator can be the contrast ratio, ranging from 0 to 1. The plurality of visibility thresholds can be [0, 0.2, 0.4, 0.75]. The ranges of values ​​can then include a minimum value range [0, 0.2[, a maximum value range [0.75, 1], and two intermediate value ranges [0.2, 0.4[ and [0.4, 0.75[, each of the value ranges being associated with a distinct respective color, for example red, orange, yellow, and blue.

[0030] Alternatively, the method may include determining a color gradient, and, for each of the selected pixels, displaying the selected pixel in a gradual color according to the visibility level of the selected pixel.

[0031] For example, the method may include determining a color gradient from red to green, in order to gradually represent low to high visibility indicator values.

[0032] In one embodiment, the visibility threshold can be chosen from: a threshold defined by a standard, such as ISO 15008, a minimum visibility threshold to compensate for a vision defect or partial impairment of sight due to aging; or a minimum visibility threshold determined to ensure satisfactory visibility for at least a predetermined percentage of a population. Initial image

[0033] Obtaining the initial image may include a digital simulation and / or a shot of the screen when it displays the visual information.

[0034] By "obtaining an image including a digital simulation" is meant any computer or digital method enabling the generation of an image, whether entirely, from digital data or models, or by modification, transformation or partial processing of a pre-existing image or photograph.

[0035] "Screen shooting" means any method for capturing a scene or object in the form of an image, for example a photograph, a video, or any other method of visual acquisition carried out using an optical, electronic or digital device, in particular a camera.

[0036] In one embodiment, obtaining the initial image includes a digital simulation of the screen as it displays the visual information. Preferably, the digital simulation is configured to generate an image of the screen as the initial image. The initial image thus generated by the digital simulation can be used to implement the steps of the process according to the invention. In another embodiment, obtaining the initial image includes a photograph of the screen as it displays the visual information.

[0037] In one embodiment, obtaining the initial image involves obtaining a preliminary image from a shot of the screen when it displays the visual information, in particular a photograph, the initial image being obtained by a digital simulation configured to modify at least partially the preliminary image into the initial image.

[0038] Preferably, the resolution of the initial image is equal to the resolution of the analyzed physical screen.

[0039] The preliminary image may have a preliminary resolution different from the screen resolution, the numerical simulation then involving a resizing of the preliminary image.

[0040] In particular, when the preliminary resolution is greater than the screen resolution, the resizing then involves a reduction in the definition and / or the preliminary resolution.

[0041] Image resolution refers to the number of pixels contained in the image per unit length.

[0042] The definition of an image refers to the number of pixels that an image contains in width and height, i.e. the number of columns and number of rows of pixels.

[0043] In one embodiment, the resizing may involve increasing the preliminary definition and / or resolution by implementing a "super-resolution" type algorithm. "Super-resolution" here refers to a set of algorithms and techniques used to improve, increase the resolution of a given image, or even upsample it. "Super-resolution" type algorithms may include, in particular, classical interpolation methods, machine learning methods, or even deep learning methods. The use of super-resolution algorithms can advantageously be combined with a restricted selection of pixels from the initial image, in other words, filtering, in order to denoise the resized image.

[0044] The initial image may be representative of a user's point of view observing the screen.

[0045] By "viewpoint of a user observing the screen," it is understood that the user is in a specific position and observes the screen from that position. In particular, the viewpoint may vary depending on the user's distance from the screen, and / or at least one angle between a direction normal to the screen and a line of sight from the user to the screen. The viewpoint may also vary depending on the user's height.

[0046] By "representative of a viewpoint of a user observing the screen", it is necessary to understand, for example, that the initial image is a visual representation of a perspective, in other words of a field of vision, or of a view of the user when he looks at the screen according to the viewpoint.

[0047] In one embodiment, the screen is a display panel in a motor vehicle, and the initial image is representative of the viewpoint of the user observing the screen from inside the passenger compartment of the motor vehicle, for example from the driver's seat.

[0048] The initial image may represent one or more reflections of one or more objects in the environment on the screen.

[0049] The term “camera” here encompasses all digital image acquisition devices operating in the visible or near-infrared range, of any resolution and technology.

[0050] The term "environment" refers both to any object outside the vehicle and to objects inside the vehicle, particularly within the vehicle's passenger compartment. Generally speaking, any source of reflection on the screen is considered an object. For example, this could be the reflection on the screen of clothing worn by a user near the screen, or of a light source.

[0051] In one embodiment, obtaining the initial image includes a digital simulation of the screen when it displays the visual information and when it is subjected to lighting.

[0052] In another embodiment, obtaining the initial image includes taking a picture of the screen when it displays the visual information and when it is subjected to lighting.

[0053] The term "lighting" refers to any light source, whether natural or artificial, that may influence the visibility of the screen. For example, lighting could be sunlight, a vehicle's headlight, a streetlamp, a nearby illuminated display system, or a reflection from a shiny surface. As another example, when the screen is a vehicle's dashboard, the lighting could be the vehicle's interior lighting.

[0054] The invention can thus make it possible, depending on the lighting, to identify areas of a vehicle's screen with a high visibility indicator, in other words suitable for displaying pictograms essential to driving the vehicle, and areas with a lower visibility indicator that can thus be used to display less essential visual information. Image to luminance conversion

[0055] The conversion of the initial image into the luminance image may include, for each pixel of the initial image, the determination of a luminance calculated according to a weighted average of the RGB values, in other words the values ​​in a red green blue encoding, of the corresponding pixel of the initial image.

[0056] The "RGB" (red-green-blue) value of a pixel refers to a color model in which each color is represented by a mixture of three principal components: Red (R), Green (G), and Blue (B). Each component takes, for example, a value between 0 and 255, thus determining the intensity of the respective component in the pixel.

[0057] In particular, the luminance Lum; of a pixel i can be expressed as

[0058] Lumj = + ^G,- + w^, with wB <wR < wG

[0059] ,Ri, G; and B; being respectively the values ​​of the red, green and blue components, wR , wG , wb being respectively weighting coefficients respectively associated with the colors red, green, and blue, the sum of the weighting coefficients being equal to 1.

[0060] For example, wR = 0.212671, wG = 0.715160 and wB = 0.072169.

[0061] Alternatively, the image-to-luminance conversion may include, for each pixel of the initial image, the determination of a luminance calculated according to a weighted average of the colorimetric components of another colorimetric system, for example HSV, weighted so as to increase the contribution of green relative to the contribution of blue in the calculated luminance.

[0062] In general, the luminance of a pixel can be expressed as a weighted sum, according to weights, of color components, where the weights are chosen so as to accentuate, for example, the contribution of green relative to the contribution of red and / or blue, and / or accentuate the contribution of red relative to the contribution of blue in the calculated luminance. Alternatively, other expressions for luminance are possible. Visibility level indicator

[0063] By "visibility indicator of said pixel", otherwise called indicator of a level of readability, or contrast indicator, we mean for example a quantity representative of luminance differences of pixels around said pixel.

[0064] The visibility indicator may be a quantity representing the differences and / or luminances of other pixels, the other pixels being determined from the position of said pixel in the luminance image. For example, the visibility indicator of said pixel may be a weighted average of the luminances of the other pixels. As a further example, the other pixels may comprise a first group of pixels and a second group of pixels, the visibility indicator being a ratio between a first geometric mean of the luminances of the first group of pixels and a second geometric mean of the luminances of the second group of pixels.

[0065] In particular, the other pixels can be determined from a neighborhood threshold, the other pixels being the pixels located at a distance of said pixel, in number of pixels, less than the neighborhood threshold.

[0066] Preferably, the visibility indicator is a numerical quantity.

[0067] Preferably, the visibility indicator has a value in the set of numbers real numbers, especially in the set of positive real numbers.

[0068] The determination of the visibility indicator of said pixel may include the determination of a minimum neighboring luminance and a neighboring luminance maximum, determined among the luminances of the neighboring pixels of said pixel, the indicator of the visibility level of said pixel being representative of the ratio between the maximum neighboring luminance and the minimum neighboring luminance.

[0069] By "neighboring pixels of said pixel," we refer here to contiguous pixels, that is to say, adjacent, in other words, the closest pixels to said pixel. A pixel of an image may have between three and eight neighboring pixels, in particular three neighboring pixels if said pixel is in a corner of the image, five pixels if said pixel is on an edge of the image excluding one of the corners of the image, and eight pixels in the remaining case.

[0070] In particular, the visibility indicator of said pixel can be chosen from the contrast ratio, the contrast, and the visibility level.

[0071] In one embodiment, the visibility indicator of said pixel is the contrast ratio (also called "contrast ratio" in English), the contrast ratio TCx; of a pixel i, expressed as

[0072] Liim_vààne_max. TCXj — voisîfjg mjn. - - L

[0073] , Lum_voisine_max / being the maximum neighboring luminance of pixel i, and Lum_voisine_min t being the minimum neighboring luminance of pixel i.

[0074] In one embodiment, the visibility indicator of said pixel is the contrast, the contrast Cx; of a pixel i expressed as

[0075] Cx^TCXt-l

[0076] TCxi being the contrast ratio of pixel i.

[0077] In one embodiment, the visibility indicator is the visibility level, the visibility level of a pixel i being expressed as

[0078] VT =_^L_

[0079] Cxi being the contrast of pixel i, and Cseuil being a reference contrast value.

[0080] The step of determining the visibility indicator makes it possible to have a local note of contrast ratio, and / or contrast, and / or visibility level, at the pixel level.

[0081] In one embodiment, the method comprises determining, for at least one area, i.e., a set of pixels, an average of the specified contrast ratios, contrast, or visibility levels. In particular, the invention can thus make it possible to obtain an overall score for the contrast ratio, contrast, or visibility level over the entire area, i.e., the set of pixels, or even over the entire screen. Edge detection

[0082] At least part of the contours may correspond to boundaries between background pixels of the screen and pixels of the visual information, the other part of the contours being notably related to lighting, the lighting producing unwanted reflections on the screen. Edge detection can thus make it possible to reveal elements related to the visual information, and / or therefore to denoise the displayed image.

[0083] The edge detection step (c) may include the implementation of an edge detection algorithm, the edge detection algorithm being a Sobel filter, otherwise known as the Sobel operator or the Sobel-Feldman operator.

[0084] The term "edge detection algorithm" here refers to any technique used in image processing and computer vision to identify edges or contours, in other words, iso-contrast curves, in an image. These contours represent the boundaries between different regions of the image where there are significant changes in visual properties such as intensity, color, or texture.

[0085] Alternatively, the edge detection algorithm can be chosen from any other algorithm known to a person skilled in the art. For example, the edge detection algorithm can be chosen from a convolutional neural network, a Canny filter, a Prewitt operator, Laplacian-based detection (or LoG, from the English "Laplacian of Gaussian"), a Roberts algorithm, or a combination thereof.

[0086] Preferably, edge detection is performed along two orthogonal directions of the luminance image. The visual information containing edges along a horizontal direction and a vertical direction, thus the edge detection is more precise. Screen

[0087] The term “screen” refers, for example, to any electronic display system configured to present one or more visual information, static or dynamic, such as text, images, graphics, videos or any other visual content.

[0088] The screen can be a screen of a human-machine interface (or "HMI").

[0089] The screen may be a screen of a human-machine interface present in a vehicle passenger compartment.

[0090] For example, the screen can be a display panel of a motor vehicle, an aircraft or a marine vehicle.

[0091] Alternatively, the screen may be a screen of an electronic device, in particular a mobile phone, a screen of a household appliance, or a display screen of a public infrastructure, for example a notice board near a train station, or a display board showing the number of available parking spaces.

[0092] Alternatively, the screen can be a paper poster displaying graphic content, which can be analyzed in the form of a luminance-calibrated HDR photo.

[0093] The screen may be polygonal in shape, in particular rectangular, circular, or other.

[0094] The screen may be substantially flat over its entire surface. Alternatively, the screen may be curved over at least part of its surface, or even over its entire surface.

[0095] The screen can be of any type, for example an LCD (from the English “Liquid Crystal Display”), OLED (or “Organic Light-Emitting Diode”), or AMOLED (or “Active Matrix Organic Light-Emitting Diode”).

[0096] The screen resolution can be high definition (HD) i.e. 1280 pixels by 720 pixels, Full HD i.e. 1920 pixels by 1080 pixels, 4K i.e. 3840 pixels by 2160 pixels, or higher. Computer Program Product

[0097] The invention also relates, independently or in combination with the above, to a computer program comprising instructions, executable by a microprocessor or a microcontroller, for the implementation of the process as defined above, when executed by the microprocessor or the microcontroller.

[0098] The invention also relates, independently or in combination with the above, to an electronic device, configured to implement the steps of the process defined above. Brief description of the drawings

[0099] [Fig.1] Fig.1 illustrates steps of an example of a process according to the invention.

[0100] [Fig.2] The [Fig.2] is a schematic representation of the initial image.

[0101] [Fig.3] Fig.3 illustrates, schematically and partially, the detection of contours of the initial image of [Fig.2].

[0102] [Fig.4] Fig.4 illustrates, schematically and partially, the determination of the visibility indicator analysis area;

[0103] [Fig.5] The [Fig.5] is a schematic representation of the displayed and filtered image.

[0104] [Fig.6] Fig.6 is an example of visibility levels for display with color levels.

[0105] [Fig.7] Fig.7 is a schematic representation of the image displayed with color levels by visibility level.

[0106] [Fig.8] The [Fig.8] is a filtered schematic representation of the [Fig.7].

[0107] [Fig.9] [Fig.9] illustrates, schematically, a detail of [Fig.8], and the spatial representation of the invention Detailed description

[0108] An example of implementation of process 1 according to the invention is described with reference to [Fig. 1],

[0109] The process 1 includes a step 10 of obtaining the initial image 11 when the screen displays a plurality of visual information visible on [Fig.2] in 12, 13, 14. Note that [Fig.2] was initially a color image.

[0110] The initial image comprises a digital simulation of the screen when it displays the plurality of visual information in a particular lighting context, which results in differences in brightness on the screen, with areas that are very bright because they are directly exposed to light (16) or exposed to reflections (17) of light from elements in the screen's environment, and less brightly lit areas (18) where the light source is masked. As illustrated, the screen can be a display panel in a motor vehicle, representative of the viewpoint of the user observing the screen, for example from the driver's seat, in particular with areas that are directly illuminated (16) and areas that are masked depending on the position of the light source, and / or the layout of the screen and / or the elements surrounding the screen, the latter being able to cause situations of masking or reflections on the screen.

[0111] In this example, the initial image 11 is generated entirely digitally. In particular, the initial image 11 may include several reflections 17 of one or more objects from the environment on the screen.

[0112] The process 1 includes a step 20 of converting the initial image 11 into a luminance image.

[0113] The conversion of the initial image 11 into the luminance image may include, for each pixel of the initial image 11, determining a luminance calculated as a weighted average of the RGB values ​​of the corresponding pixel of the initial image 11, the luminance Lum of the pixel being expressed as

[0114] Lum = WgB + wgG + wbB

[0115] R, G and B being respectively the values ​​of the red, green and blue components of the pixel, wR, wG, wb being respectively weighting coefficients respectively associated with the colors red, green, and blue, the sum of the weighting coefficients being equal to 1.

[0116] Optionally, step 20 may include an intermediate display of a luminance representation of the image, the luminance values ​​of the pixels of the initial image 11 being represented according to a color gradient.

[0117] The method 1 may include a step 30 of edge detection of the luminance image, the edges comprising a plurality of pixels, including the implementation of an edge detection algorithm, for example a Sobel filter.

[0118] Fig. 3 is a representation 31 of the luminance image after the edge detection step 32. As illustrated, the representation 31 has a plurality of edges 32, otherwise called iso-contrast curves, shown in white.

[0119] The method 1 may include a step 40 of determining, for each of the pixels of the plurality of pixels, the contrast ratio of said pixel.

[0120] Alternatively, the visibility indicator is the visibility level, the visibility level of a PO pixel being expressed as

[0121] VL=7f^, Cseuil

[0122] Cx being the contrast of pixel PO, and Cseuil being a reference contrast value. For example, Cseuil may be a reference contrast value of a normalized Blackwell or Adrian model.

[0123] The bottom of [Fig.4] is a representation of the image after contour detection 32, in particular enlarged on a textual visual information 13 indicating a speed of the vehicle in km / h, and on which is indicated the position of a pixel of interest PO on one of the contour curves.

[0124] The top of [Fig. 4] is a representation of the luminance image, shown on a greyscale. On this image are indicated the positions of two pixels of interest PI and P2, adjacent to the position of pixel PO.

[0125] As can be seen in [Fig. 4], the determination of the visibility indicator of pixel PO includes the determination of a minimum neighboring luminance PI and a maximum neighboring luminance P2, determined among the luminances of the neighboring pixels of said pixel PO on the luminance image representation, the visibility level indicator of said pixel being representative of the ratio between the maximum neighboring luminance and the minimum neighboring luminance, the pixel PO being included in a contour 32. In other words, the pixel PO is identified on the contours, and, from its corresponding position in the luminance image representation, the luminance values ​​of the neighboring pixels can be determined, in particular in order to identify the pixel of maximum neighboring luminance (P2) and minimum neighboring luminance (PI) and to determine the visibility indicator of the pixel PO.

[0126] Pixels PI and P2 belong respectively to the background of the screen, and to the useful display of the screen, in particular of the visual information 12. They are adjacent to pixel PO.

[0127] It is thus possible, in particular, to automatically determine the separations between background pixels and pixels of graphic content or background noise related to the lighting context, by taking contrast measurements only on the points of interest represented by the contour curves. The preceding steps 10, 20, 30, 40 can thus be used to determine the differences in contrast ratio, or visibility levels, of the various visual information displayed on the screen.

[0128] The method may include a step 50 of displaying an image 50 representative of the visibility level indicator, for example the visibility level, for each of the pixels of a selection of pixels from the plurality of pixels.

[0129] As illustrated in [Fig. 5], the pixel selection may include pixels whose visibility indicator exceeds a predetermined filtering threshold. Such a display allows for the instantaneous visual identification of visual information exceeding the filtering threshold, and for quick and easy validation of the display's compliance with predefined requirements or applicable regulations.

[0130] The method may include determining at least one visibility threshold, and for each of the filtered, in other words selected, pixels, displaying the selected pixel in a first color when the visibility indicator of the selected pixel is strictly less than the visibility threshold, and displaying the selected pixel in a second color when the visibility indicator of the selected pixel is greater than the visibility threshold.

[0131] The display step 50 is thus enriched by colour information, which can allow for the efficient visualization of areas of the screen that meet predefined requirements, or regulations in force.

[0132] Figure 8 is a representation of the displayed image 50, filtered to isolate the graphic elements. In this example, the pixel selection includes only the pixels from the plurality of pixels with a visibility indicator, here the level of Visibility levels must exceed the predefined filtering threshold. Therefore, elements with visibility levels below the filtering threshold are not displayed. A color gradient, not visible in the black and white figure, is used to represent the different visibility levels, allowing for immediate visualization of the displays with the best visibility.

[0133] Figure 9 illustrates a detail 90 of the textual visual information 13 of Figure 4. As illustrated, the invention makes it possible to identify different levels of visibility indicator for a pictogram. It appears in Figure 9 that the contrast of the right-hand side 94 of the visual information 13 is less strong than the contrast of the left-hand side 92. This figure illustrates the spatial calculation of the contrast criterion, contrast ratio, or visibility level, highlighting the fact that the pictogram does not meet the visibility criterion across its entire graphic representation on the screen, particularly not on its right-hand side 94.

[0134] Fig. 7 is another representation of the displayed image 50, with a visibility threshold and two display colours of the selected pixels, the selection this time comprising the plurality of pixels.

[0135] The method may include determining a plurality of visibility thresholds defining disjoint ranges of values, the plurality of visibility thresholds comprising a minimum visibility threshold and a maximum visibility threshold, the ranges of values ​​covering all the values ​​of the visibility indicators of the selected pixels, the ranges of values ​​comprising: -a range of minimum values ​​comprising a set of values ​​below the minimum visibility threshold, -a range of maximum values ​​including a set of values ​​exceeding the maximum visibility threshold, and -at least one range of intermediate values ​​comprising a set of values ​​between two consecutive visibility thresholds,

[0136] the ranges of values ​​being associated with distinct respective colours,

[0137] the method comprising, for each of the selected pixels, displaying the selected pixel in the color associated with the range of values ​​including the visibility indicator of said pixel

[0138] Alternatively, the method may include determining one or more color gradients, and, for each of the selected pixels, displaying the selected pixel in a gradual color, defined by one of the color gradients, according to the visibility level of the selected pixel.

[0139] Fig. 6 is a table 60 illustrating visibility thresholds. In this example, the visibility thresholds are visibility level thresholds 61.

[0140] When the visibility indicator is the visibility level, the plurality of visibility thresholds can be chosen from among the visibility level thresholds 61.

[0141] For each visibility level threshold 61, table 60 includes a description 62 of said threshold 61, a representation 63 of the color associated with said threshold 61, and an RGB encoding 64 of the color associated with said threshold 61.

[0142] In this example, the plurality of visibility level thresholds 61 comprises six thresholds, defining disjoint ranges of values.

[0143] As an example, the value ranges include the maximum value range comprising all visibility levels above 15, exhibiting relatively satisfactory performance within the meaning of ISO 15008.

[0144] Figures 10 to 14 illustrate another example of implementation of method 1 according to the invention, from another initial image 11' digitally simulated.

[0145] The invention allows digitally to visualize instantly and precisely the visibility or contrast levels of all the pixels of the HMI of a vehicle, as well as to vary the visibility thresholds, in order to be able to instantly, during the vehicle design phase, determine the visibility levels of the HMI displays according to various lighting situations.

[0146] Thus, it is possible to distribute the arrangement of visual information, for example pictograms, on the screen in the best possible way, taking into account that some visual information is essential and needs to be arranged so that its visibility is optimal in all conditions, relative to other less priority visual information which can thus be placed in areas where visibility is less good.

[0147] Also, it is possible to detect the elements of the screen environment that cause reflections, and thus to move these elements, and / or modify their structure, and / or validate or invalidate a screen during a phase upstream of the vehicle design.

[0148] Furthermore, the computer program configured to implement the steps of the method of the invention can allow for real-time variation of the visibility thresholds and the filtering threshold, in order to instantly display the screen elements that meet the visibility requirements, and conversely, those that do not. The computer program and its use form an integral part of the invention.

Claims

Demands

1. A method (1) for determining a visibility indicator of visual information (12, 13; 14) on a screen, when the visual information (12, 13; 14) is displayed by the screen, characterized in that it comprises the following steps: a. Obtaining (10) an initial image (11) of the screen when the screen displays the visual information (12, 13, 14), b. Converting (20) the initial image (11) into a luminance image, c. Detecting (30) edges (32) of the luminance image, the edges (32) comprising a plurality of pixels (PO), d. Determination (40), for each of the pixels (PO) of the plurality of pixels, of a visibility indicator of said pixel (PO), by comparison of the luminances of other pixels (P1,P2), the other pixels (P1,P2) being determined from a position of said pixel (PO) in the luminance image.

2. A method according to the preceding claim, comprising a step (e) of displaying (50) an image representative of the visibility level indicator for each of the pixels of a selection of pixels from the plurality of pixels.

3. Method according to claim 2, pixel selection comprising only those pixels whose visibility indicator is greater than a predetermined filtering threshold.

4. A method according to any one of the two preceding claims, comprising determining at least one visibility threshold, and for each of the selected pixels, displaying the selected pixel in a first color when the visibility indicator of the selected pixel is strictly less than the visibility threshold, and displaying the selected pixel in a second color otherwise.

5. A method according to any one of the preceding claims, wherein obtaining (10) the initial image (11) includes a digital simulation and / or a shot, of the screen when it displays the visual information (12,13,14).

6. A method according to any one of the preceding claims, wherein the initial image (11) is representative of a viewpoint of a user observing the screen.

7. A method according to any one of the preceding claims, wherein the conversion (20) of the initial image into the luminance image comprises, for each pixel (PO) of the initial image (11), the determination of a luminance calculated as a weighted average of the RGB values ​​of the corresponding pixel (PO) of the initial image (11), the luminance Lum; of a pixel i being expressed as Lumi = wRJti + w(i.Gi + wBBi, with wB <wR <wG ,Ri, G; et B; étant respectivement les valeurs des composantes rouge, vert et bleu, wR, wG, wB étant respectivement des coefficients de pondération respectivement associés aux couleurs rouge, verte, et bleu, la somme des coefficients de pondération étant égale à 1.

8. Method according to the preceding claim, the determination of the visibility indicator of said pixel (PO) comprises the determination of a minimum neighboring luminance and a maximum neighboring luminance, determined among the luminances of the neighboring pixels (P1,P2) of said pixel, the indicator of the visibility level of said pixel (PO) being representative of the ratio between the maximum neighboring luminance and the minimum neighboring luminance.

9. A method according to any one of the preceding claims, wherein the screen is a screen of a human-machine interface present in a vehicle cabin.

10. Computer program comprising instructions, executable by a microprocessor or microcontroller, for implementing the method according to any one of the preceding claims, when executed by the microprocessor or microcontroller.

11. Electronic device, configured to carry out the steps of the process according to any one of claims 1 to 10.

Images