Method and system for characterizing keratin fibers, in particular human eyelashes.

Abstract

Method for characterizing eyelashes or eyebrows The present invention relates to a computer-implemented method for characterizing eyelashes or eyebrows, comprising the following steps: - receiving data corresponding to a set of pixels of a close-up image of a body area comprising a plurality of eyelashes, preferably the entirety of an eyelash fringe or an eyebrow, to be characterized; - applying at least one computer vision step so as to obtain, from the received image data, a numerical evaluation of at least one characteristic numerical parameter among, in particular, the number of fibers, an average length of the fibers, an average thickness of the fibers, a curvature of the fibers. Figure for abstract: Fig 1

Description

Title of the invention: Method and system for characterizing keratin fibers, in particular human eyelashes.

[0001] The present invention relates to a method for characterizing keratin fibers such as eyelashes or eyebrows, with a view to selecting and applying a cosmetic composition to said keratin fibers.

[0002] By “cosmetic products” is meant any product as defined in Regulation (EC) No 1223 / 2009 of the European Parliament and of the Council of 30 November 2009, relating to cosmetic products.

[0003] Thus, a cosmetic product is generally defined as being a substance or mixture intended to be placed in contact with superficial parts of the human body (epidermis, hair and capillary systems, nails, lips and external genital organs) or with the teeth and oral mucous membranes with a view, exclusively or mainly, to cleaning them, perfuming them, modifying their appearance, protecting them, keeping them in good condition or correcting body odors.

[0004] The present invention relates more particularly to the characterization of eyelashes, in particular human eyelashes, with a view to the application of a makeup or care composition such as mascara thereon.

[0005] Mascara means a composition intended to be applied to the eyelashes. It may in particular be an eyelash makeup composition, an eyelash makeup base (or "base coat"), a composition to be applied over mascara (or "top coat"), or even a cosmetic eyelash treatment composition. Mascara is more particularly intended for human eyelashes, but also for false eyelashes.

[0006] The application of mascara aims in particular to increase the intensity of the look, in particular through a more or less significant increase in the volume and / or length of the eyelashes. The principle consists of depositing a desired quantity of material on the eyelashes so as to obtain this volumizing and / or lengthening effect.

[0007] The cosmetic product is applied by means of an applicator.

[0008] Generally, an applicator comprises an application member connected to a gripping member via a rod.

[0009] The application member defines an application surface and has a main body or core, generally elongated, capable of carrying application elements extending projecting from said core. Preferably, the application elements extend in a general direction substantially normal to the core (in particular radial).

[0010] During application, the application member is loaded with cosmetic product and brought into contact with the fibers in order to allow the product to be deposited on them. The application elements, spaced apart from each other, form cosmetic product reservoir areas. They also make it possible to separate / comb the eyelashes so as to optimize the deposit of product on each eyelash.

[0011] Mascara and applicator sets are known which are designed to provide different effects, and in particular more or less significant volumizing and / or lengthening effects depending on the composition of the mascara and the applicator used. Thus, each commercial product aims for a particular effect.

[0012] Obviously, the final result also depends on the initial characteristics of the user's eyelashes. Thus, a user with dense and thick eyelashes will only need a light mascara to obtain a marked volumizing effect, whereas a user with thin and sparse eyelashes will need to use a specific mascara that allows the eyelashes to be heavily loaded with product.

[0013] Finding the appropriate product according to the desired effects is complex and may require several product trials which can discourage the consumer.

[0014] One objective of the cosmetics industry is to always improve the experience lived by its consumers and to offer products and compositions that are ever more adapted to their needs and their specific characteristics.

[0015] Therefore, there is a need to develop systems that allow the user to better understand the characteristics of their eyelashes, in order to more easily select a product likely to give them the desired effect.

[0016] Document JP5279213B2 proposes for this purpose a tool for evaluating different parameters of eyelashes such as their length, their density and their curvature. Each parameter is evaluated manually by reference to a corresponding ruler marked on the tool.

[0017] However, such a tool is not easy to use, requires several positioning and reading operations, and severely limits the possibilities of interaction with the user and the consumer.

[0018] In order to address these limitations, the present invention aims at a method for characterizing eyelashes comprising the following steps aimed at:

[0019] - receive data corresponding to a set of pixels forming an image close-up of a body area comprising a plurality of eyelashes, preferably an entire fringe of eyelashes, to be characterized;

[0020] - applying at least one computer vision step to the received image data And,

[0021] - obtaining from the computer vision step, a numerical evaluation of at least a characteristic parameter of said eyelashes, including, in particular, a number of fibers, an average length of the fibers, an average thickness of the fibers.

[0022] It has indeed been unexpectedly noted that the use of “machine vision” techniques could allow the quantitative characterization of eyelashes. Eyelashes are in fact small objects generally having a diameter of the order of a hundred micrometers in diameter and relatively few in number (of the order of a hundred for the eyelashes of an upper fringe of the eye). Thus, their identification and processing by computer vision techniques could appear particularly difficult.

[0023] Indeed, it is known to use computer vision techniques to detect the iris of an eye in an image (see for example, Jus Lozej, Blaz Meden, Vitomir Strucy, Peter Peer, "End-to-End Iris Segmentation using U-Net", 2018 IEEE International Work Conference on Bioinspired Intelligence (IWOBI), DOI:10.1109 / IWOBI.2018.8464213). In such an iris detection method, the eyelashes must be eliminated from the image, i.e. the eyelash pixels in the image are classified as not belonging to the object being searched for, and little attention is paid to their effective identification and segmentation.

[0024] In other words, the present method aims to reverse the method by choosing to select and retain the eyelash pixels on an image in order to allow the extraction and determination of characteristic digital parameters. It is thus notably possible to implement a U-Net network as used for the iris and to train it appropriately for eyelash recognition.

[0025] Furthermore, such a method allows the evaluation of several parameters from the same image, where appropriate by carrying out several image analysis and / or calculation steps after the image has been subjected to the computer vision step.

[0026] It is obviously possible to submit several images, taken in different ways (in profile, closed eye, etc.), to the computer vision steps so as to reinforce the reliability of the evaluation (averaging of the results, etc.) or even to evaluate other parameters which can be accessed more reliably from an image taken from another angle (for example, for the evaluation of the curvature, an image of the eyelashes seen from the side may be preferred).

[0027] By "close-up image" is meant a framing which isolates a part of the human body, in this case, the close-up image is centered on an area of ​​the eye comprising at least the eyelashes. Preferably, the image also includes the eye, in particular its iris. In a complementary manner, the image may also include the eyebrows corresponding to the area of ​​the eye considered. Preferably, the close-up image excludes the nose, in particular a wing of the nose.

[0028] Using a close-up image makes it possible to best ensure that the eyelashes present in the image will occupy at least a few pixels in width.

[0029] The method may comprise a prior step of image acquisition by a camera, the image acquisition step being able to be advantageously preceded by a step of removing makeup from the eyelashes and / or a step of combing the eyelashes, the combing preferably being carried out after removing makeup.

[0030] These steps, and in particular the combing step, make it possible to optimize the separation and individualization of the eyelashes so as to facilitate their rendering and identification on the image taken.

[0031] The acquisition step is carried out with a digital camera, in particular integrated into a tablet or a personal telephone, comprising a sensor of at least 6 Mpx, preferably at least 8 Mpx, or even at least 12 Mpx. The image can advantageously be acquired with an HDR (“High Dynamic Range”) mode. The close-up image preferably has a minimum resolution of 4K, better still 8K and contains at least 8 Mpx, preferably 12 Mpx or even 24 Mpx.

[0032] Such resolutions make it possible to ensure that most of the eyelash fibers occupy in their diameter several pixels of width of the acquired image.

[0033] The close-up image is preferably taken directly from close up (proxiphotography) during the acquisition step, for example at a distance from the eyelashes of between 10 and 50 cm, preferably between 15 and 25 cm. Even if this is not desirable, it is obviously possible to use a zoom, preferably optical, but a digital zoom can also be used provided that the image quality and resolution remain sufficient. Preferably, a zoom is not used. A “macro” mode of the camera can also be used.

[0034] However, the close-up image can also result from a cropping of a larger image, in particular a so-called “full-face” image. The cropping on the area of ​​interest, in particular the eye and eyebrow area, can itself be carried out by an image processing step making it possible to recognize and segment elements of the face.

[0035] To this end, it will be possible to implement techniques for detecting and segmenting facial features, such as those described for example in the document Zakia Hammal, NicolasEveno, AliceCaplier, Pierre-YvesCoulon, “Parametric models for facial features segmentation”, SignalProcessing, Elsevier, 2005, 86, pp.399-413. hal-00121793; or software modules such as face_recognition available at https: / / github.com / ageitgey / face_recognition. The identification and detection of the eye and / or eyebrows thus makes it possible to perform appropriate cropping of the image in order to obtain the desired close-up image.

[0036] Depending on the quality of the image used, the eyelashes may represent only about 0.5% of the total number of pixels (about 60,000 eyelash pixels for an image taken with a 12 million pixel sensor). There is therefore a need to optimize their detection in the context of the method of the present application.

[0037] Advantageously, the image acquisition step comprises a step of focusing the camera on the eyelashes. In particular, an autofocus, or automatic focusing, can be carried out by an operator before taking the image. This makes it possible to optimize the sharpness of the eyelashes in the image.

[0038] Alternatively or in a complementary manner, the acquisition step comprises a step of verifying the sharpness of the image, in particular by applying an algorithmic criterion such as a variance of a Laplacian (cf. R. Bansal, G. Raj and T. Choudhury, "Blur image detection using Laplacian operator and Open-CV," 2016 International Conference System Modeling & Advancement in Research Trends (SMART), 2016, pp. 63-67, doi: 10.1109 / SYSMART.2016.7894491). Thus, an image presenting insufficient sharpness compared to a predefined threshold could be rejected and / or the operator could be asked if he wishes to take a new image. A degree of sharpness can also be presented directly to an operator in order to assist him in the evaluation and the shooting. Flash or additional lighting can be used to advantage when taking pictures.The use of such additional lighting helps to limit the impact of variations in ambient light.

[0039] Advantageously, the image acquisition step is carried out from a low angle, preferably at an angle of between 10 and 45 degrees, preferably at an angle of between 15 and 20 degrees, below the fringe of eyelashes considered.

[0040] Advantageously, the acquisition step is also carried out with a subject preferably standing upright, and holding his head substantially straight. Very preferably, the acquisition step is carried out on a subject, eye open, and whose gaze is directed upwards.

[0041] In this way, the eyelashes or the fringe of eyelashes considered are generally substantially parallel to a plane of the camera lens, which makes it possible to optimize their detection and characterization by the computer vision step.

[0042] According to a first embodiment, the computer vision step implements a direct regression method, in particular by applying a residual neural network, in particular of the ResNet type (see, for example (K. He, X. Zhang, S. Ren and J. Sun, "Deep Residual Learning for Image Recognition," 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2016, pp. 770-778, doi: 10.1109 / CVPR.2016.90.).

[0043] The use of such a method makes it possible to obtain a direct numerical estimate of the characteristic parameters sought by the neural network from the image. Such a method can advantageously make it possible to avoid one or more additional steps of segmentation of the eyelashes.

[0044] According to a second, alternative or complementary embodiment, the computer vision step comprises at least one step of identification by segmentation of the eyelash fibers.

[0045] Advantageously, the computer vision step comprises a step of identification by segmentation of roots and / or tips of the eyelashes.

[0046] The segmentation steps can in particular be implemented by applying an artificial intelligence model trained in a corresponding manner to enable the detection of the objects sought. The segmentation steps are in particular carried out by classifying the pixels as belonging or not to the object sought and makes it possible to obtain one or more corresponding segmentation masks.

[0047] In particular, the segmentation steps are carried out by implementing U-Net neural networks (Ronneberger O., Fischer P., Brox T. (2015) U-Net: Convolutional Networks for Biomedical Image Segmentation. In: Navab N., Hornegger J., Wells W., Frangi A. (eds) Medical Image Computing and Computer-As sisted Intervention - MICCAI 2015. MICCAI 2015. Lecture Notes in Computer Science, vol 9351. Springer, Cham. https: / / doi.org / 10.1007 / 978-3-319-24574-4_28). These networks allow the image pixels to be classified into different categories; in the present case of eyelash characterization, the following categories could be used: fiber, root, tip, among others. The pixels thus classified allow the generation of one or more associated segmentation masks.

[0048] It has indeed been found that such networks are particularly suitable for the detection of fibers.

[0049] However, if the submitted image includes several types of fibers, it may be difficult for the network to disambiguate them. Thus, for example, if the close-up image includes both eyelashes and eyebrows, the implemented neural network may not distinguish them, in particular due to a local classification method not taking into account a global positioning of the pixel considered to estimate its belonging to one or other element of the face.

[0050] In order to resolve such a difficulty, it may be necessary to carry out a preliminary step of cropping the submitted close-up image in which the eyebrows are eliminated, or to apply the computer vision steps, and in particular segmentation, only to an area of ​​the reduced image where the eyelashes are present, excluding the eyebrows. This reduced area may be obtained by first implementing a step of classifying the pixels of the image as belonging to an eyelash area or not. This classification step may implement a network U-Net type neural network as before. Advantageously, it is also possible to perform iris detection or another reference element during this step, in particular to enable distance calibration as explained below.

[0051] Preferably, the method comprises a step of generating an image highlighting the elements identified by segmentation on the received image. The highlighting of the identified elements can in particular be done by specific coloring of the pixels of the image identified as belonging to a characteristic element. The generated image can in particular consist of at least one superposition filter, obtained from one or more segmentation masks, and capable of being displayed superimposed on the original image (possibly having undergone one or more first graphic modifications such as a color inversion).

[0052] In particular, the overlay filter may comprise pixels of a background color (for example black) not corresponding to any specific element identified by segmentation and pixels of a first color (for example red) corresponding to a first element identified by segmentation (for example eyelash fibers). A second filter or the same filter may comprise pixels of a second color (for example orange) corresponding to a second element identified by segmentation (for example the tips of the eyelashes). A third filter or the same filter may comprise pixels of a third color (for example green) corresponding to a third element identified by segmentation (for example the roots of the eyelashes).

[0053] In a complementary manner, the segmentation step(s) are followed by at least one step of calculating the desired characteristic parameter from the pixels classified by segmentation.

[0054] Preferably, the average length of the fibers is obtained using the root-tip distances of each fiber.

[0055] Preferably again, the number of fibers is determined by a pixel hopping method. The pixel hopping technique notably comprises the step of defining one or more transverse lines of pixels extending through all of the eyelashes acquired in the image (notably through the fringe of eyelashes), said transverse line being located between the tips and the roots previously identified. The pixels of this transverse intermediate line are traversed in a given direction and the fibers can be counted from the color variation of the pixels traversed. This determination can be made on the original image or on the modified image from the segmentation masks.

[0056] The distances may be determined and expressed in number of pixels or any distance characteristic of the sensor used for image acquisition.

[0057] Advantageously, the distances are determined with respect to at least one reference element present in the image of the received body area, in particular in the received close-up image. In particular, the reference element is a characteristic facial feature having a substantially fixed average dimension among a population.

[0058] In particular, when the close-up image includes the eye and / or its iris, the diameter of the iris, considered to have a standard average distance of approximately 10 mm, and / or a palpebral fissure, considered to have a standard average distance of approximately 3 cm, may be used.

[0059] These reference elements can be identified on the image by an appropriate segmentation step making it possible to count the number of pixels associated with the desired reference dimension. The number of pixels forming an eyelash, in length and / or in width, can thus be easily converted into an intelligible real distance which can be expressed in particular in standard units, in particular in centimeters or millimeters.

[0060] All or part of the characteristic parameters thus determined, in particular the number of eyelashes detected, as well as their average length and their average thickness can be displayed and presented to the user. Advantageously, each parameter is displayed on a gauge positioning them relative to a reference value, such as the average value for a population considered.

[0061] In a complementary manner, the method comprises an additional step of classifying the eyelashes among at least two typologies established from at least one characteristic parameter of said eyelashes, preferably from at least two parameters.

[0062] In particular, eyelashes can be classified according to the following typologies depending on their density (number) and their average length:

[0063] - short and sparse

[0064] - long and sparse

[0065] - medium long and medium dense

[0066] - short and dense

[0067] - long and dense

[0068] Relative characteristics, such as short / long and dense / sparse, are determined relative to a reference value for the characteristic under consideration. As with the display, the reference value may in particular be an average value for a given population.

[0069] In an advantageously complementary manner, the method comprises an additional step of choice by a user of a desired treatment result. In particular, in the case of eyelashes and the application of mascara, it may be requested to the user whether she is looking for “natural volume”, “intense volume”, or “extreme volume”.

[0070] A step of querying a database may then be carried out based on at least one determined characteristic parameter, and in particular based on the determined typology, and the desired treatment result, so as to determine at least one recommended cosmetic product.

[0071] The method comprises a step of presenting the products thus determined.

[0072] Thus, for a person with sparse and short eyelashes, and wanting a “ natural volume”, he may be offered a mascara providing a moderate volumizing effect.

[0073] A person who already has a large number of relatively long eyelashes, and also wants “natural volume” will be offered a mascara that provides a much lower volumizing effect.

[0074] Obviously, all or part of the steps implemented by computer can be carried out locally or by a remote server after sending the image data, the transmission of the data being able to be carried out by any known means, in particular by any means of wireless communication.

[0075] Preferably, the image acquisition and all or part of the data processing or data sending steps are carried out by the same portable device. In particular, this could be a portable telephone or tablet having an integrated camera.

[0076] Thus, the present invention also relates to a computer system for implementing a method according to the invention, comprising: - at least one means of importing and / or acquiring a close-up image, - calculation means for implementing all or part of the processing steps and in particular the computer vision steps and the digital evaluation steps, - at least one display means, configured to allow the display of the numerical evaluation results obtained.

[0077] Preferably, all or part of the calculation means capable of carrying out the computer vision and / or digital evaluation steps belong to a remote server separate from a device incorporating the image acquisition and / or import means.

[0078] Other aims, characteristics and advantages of the invention will appear on reading the following description, given solely by way of non-limiting example, and made with reference to the appended drawings in which:

[0079] [Fig-1] is an illustration of a screenshot of a portable tablet used for implementing the method of the invention and showing the results that can be obtained.

[0080] As detailed above, the present method implemented by computer aims mainly to allow the characterization of a user's eyelashes in order, in particular, to be able to recommend a suitable cosmetic product to her, according to the determined digital parameter(s).

[0081] To do this, the present method is implemented using a computer tablet comprising an integrated camera.

[0082] In a first step, a photo is taken of a close-up image 1 of an area of ​​the eye comprising the eyelashes, in particular an entire fringe of eyelashes and a corresponding eyebrow.

[0083] It should be noted that before acquisition, the user's eyelashes were removed from makeup and combed.

[0084] Furthermore, when taking the image, the camera is focused on the eyelashes so as to obtain the sharpest possible image of the eyelashes. The sharpness is here assessed visually by the operator.

[0085] The image is taken with the eye open, head straight and gaze directed upwards.

[0086] The image is taken at a distance from the eyelashes of about 15 cm, without using a zoom.

[0087] The image is also taken from a low angle at an angle of between 15 and 20 degrees, below the fringe of eyelashes considered.

[0088] All of the pixels constituting the image thus acquired are then processed by computer vision steps.

[0089] The processors equipping personal tablets generally do not have sufficient computing power, the image data will advantageously be transmitted to a remote server for processing before returning the result to the personal tablet for display.

[0090] During a first processing step, the initial image comprises a step of determining an area of ​​the eyelashes in the image, in order to eliminate the eyebrow area whose fibers are likely to be confused with the eyelashes.

[0091] Where appropriate, the method according to the present application may optionally be applied to an eyebrow according to the same principles. The image will then be taken substantially from the front, and the fringe of eyelashes eliminated from the image and / or from the treatment where appropriate.

[0092] During this step, the iris may also be segmented so as to use it as a reference element for the subsequent determination of distances.

[0093] The eyelash area thus isolated, the pixels of this area are subjected to a classification step allowing them to be identified as fiber pixel, tip pixel and / or root pixel.

[0094] As visible in [Fig.l], an image 100 of the initially acquired image 1 is displayed so as to highlight the identified elements by colorization. Thus, image 100 has undergone an inversion of the colors (the skin appearing in blue tones, the iris in white and the sclera in black) and a specific colorization of the fiber pixels (red), the tip pixels (orange), and the root pixels (green).

[0095] Thus, the different pixels of the eyelash zone identified and classified, the numerical parameters of interest are calculated.

[0096] First, the total number of eyelashes in the image is determined by a pixel hopping method, previously explained.

[0097] The number of eyelashes detected is displayed 10 and presented to the user in the form of a cursor 11 positioned on a gauge 12 relative to an average reference value 13 in a given population. In this case, a number of 82 eyelashes was determined and is considered slightly higher than the average number of eyelashes in a sample of people tested.

[0098] When counting eyelashes by pixel hopping, a thickness (number of pixels crossed) of each eyelash counted can also be determined. The number of eyelashes makes it possible to obtain an average thickness.

[0099] Thanks to the user's iris previously segmented and identified on the acquired image, it is possible to convert the thickness in pixels into thickness in absolute distance unit. Indeed, as indicated previously, it can be considered that an iris has a general average diameter in the population of 1 cm. The number of pixels constituting a diameter of the iris can thus be related to the number of pixels of thickness of the eyelash to determine the measurement in unit of length, in particular in cm or in mm.

[0100] As for the number of eyelashes, the determined value is displayed 20 and presented to the user in the form of a cursor 21 positioned on a gauge 22 relative to an average reference value 23 in a given population (preferably, the same population for which the average reference value of the number of eyelashes was determined).

[0101] In this case, in the example shown, an average eyelash thickness of 0.086 mm was thus determined, which is considered to be slightly lower (thinner eyelashes) than the average reference value.

[0102] Thirdly, the length of the eyelashes is also determined by the root-tip distances and an average value is determined in relation to the number of eyelashes counted.

[0103] More precisely, the length of the eyelashes can be determined in the following manner: i. first of all, a base line of the roots is determined, appearing in the form of an arc of pixels passing substantially in the middle of the roots of the fringe of eyelashes; ii. we then calculate a distance between this arc and each identified point.

[0104] Rather than an average length of eyelashes, one may choose to retain the greatest length determined as representative of the general length of eyelashes.

[0105] As for the thickness of the eyelashes, the length can be expressed in cm or mm with reference to the diameter of the iris serving as a reference element.

[0106] As for the number of eyelashes, the determined value is displayed 30 and presented to the user in the form of a cursor 31 positioned on a gauge 32 relative to an average reference value 33 in a given population (preferably, the same population for which the average reference values ​​of the number of eyelashes and the thickness of the eyelashes were determined).

[0107] In this case, in the example shown, an average eyelash length of 4.349 mm was thus determined, which is considered to be substantially lower (shorter eyelashes) than the average reference value.

[0108] Other numerical parameters characterizing the eyelashes may, if necessary, be determined by similar techniques. In particular, it will also be possible to determine a curvature of the eyelashes. In such a case, it may be preferable to acquire a new image taken in profile with respect to the eyelashes.

[0109] The characteristic parameters of the eyelashes having thus been determined, a typology of the user's eyelashes can be determined from among several predefined typologies.

[0110] Thus, in this case, the user taken as an example will belong to a typology of “medium dense and medium long” eyelashes.

[0111] A database is then queried associating products with the effect provided according to their application on certain types of eyelashes.

[0112] In this case, the user is presented with a product (PR) allowing substantial lengthening of the eyelashes while moderately increasing their volume.

Claims

Claims

1. A computer-implemented method for characterizing eyelashes or eyebrows, comprising the following steps: - receiving data corresponding to a set of pixels of a close-up image (1) of a body area comprising a plurality of eyelashes, preferably the entirety of an eyelash fringe or an eyebrow, to be characterized;- applying at least one computer vision step so as to obtain, from the received image data, a numerical evaluation (11, 21, 31) of at least one characteristic numerical parameter among, in particular, the number of fibers (10), an average length of the fibers (20), an average thickness of the fibers (30), a curvature of the fibers, said method being characterized in that it comprises the following additional steps: - identifying on the image of the received body area, a reference element having a substantially fixed average dimension among a population, - determining a number of pixels associated with the dimension of the reference element, - converting determined distances into a number of pixels in a standard unit.;

2. Method according to claim 1, characterized in that the computer vision step implements a regression method by applying a residual neural network, in particular of the ResNet type.

3. Method according to any one of claims 1 or 2, characterized in that the computer vision step comprises at least one step of identification by segmentation of the fibers of the eyelashes or eyebrows.

4. Method according to any one of claims 1 to 3, characterized in that the computer vision step comprises a step of identification by segmentation of roots and / or tips of eyelashes or eyebrows.

5. Method according to any one of claims 3 and 4, characterized in that it comprises a step of generating an image (100) highlighting the elements identified by segmentation on the received image.

6. Method according to any one of claims 3 to 5, characterized in that the segmentation step is followed by at least one step of calculating the parameter (10, 20, 30) from the pixels classified by segmentation.

7. Method according to claim 6, characterized in that it comprises the additional steps aimed at: - determining a length of the fibers by determining a number of root-tip pixels of each fiber. - determining an average length from a number of fibers counted.

8. Method according to any one of claims 3 to 7, characterized in that the number of fibers is determined by a pixel hopping method.

9. Method according to any one of claims 1 to 8, characterized in that it comprises an additional step of classifying the eyelashes or eyebrows among at least two typologies established from at least one characteristic parameter of said eyelashes, preferably from at least two parameters.

10. Method according to any one of claims 1 to 9, characterized in that it comprises a step of interrogating a database from at least one determined characteristic parameter, and in particular from the determined typology, and from a desired treatment result of the eyelashes or eyebrows chosen by a user, so as to determine at least one recommended cosmetic product (PR).

11. Method according to claim 10, characterized in that it comprises a step of presenting on a screen the recommended product(s) (PR).

Images