Computer implemented method for detecting and optionally measuring at least one protuberance

Abstract

The present invention relates to a computer-implemented method for processing data generated by scanning a surface of a person's skin or part of such surface where at least two different modules have been involved in the capture. A numerical model representing the scanned skin surface is generated from the scanned data obtained from the different modules. The joint of the data generated by one module and the at least another module gives rise to discontinuities in the surface that are not easily identifiable, for example when the surface is represented mainly by a cloud of points or by a plurality of lines, in both cases, a discrete representation of the surface. The method provides a solution in which the discontinuity is first identified and then one surface is shifted with respect to the other in a specific way resulting in a continuous surface that maintains all the shape features of the original surfaces.

Description

[0001]COMPUTER IMPLEMENTED METHOD FOR DETECTING AND OPTIONALLY MEASURING AT LEAST ONE PROTUBERANCE DESCRIPTIONFIELD OF THE INVENTIONThe present invention relates to a computer-implemented method for processing data generated by scanning a surface of a person's skin or part of such surface where at least two different modules have been involved in the capture. A numerical modelrepresenting the scanned skin surface is generated from the scanned data obtainedfrom the different modules. The joint of the data generated by one module and the atleast another module gives rise to discontinuities in the surface that are not easilyidentifiable, for example when the surface is represented mainly by a cloud of pointsor by a plurality of lines, in both cases, a discrete representation of the surface.The method provides a solution in which the discontinuity is first identified and then one surface is shifted with respect to the other in a specific way resulting in a continuous surface that maintains all the shape features of the original surfaces. PRIOR ART One area of the technique of great interest is the modeling of the shape of surfaces bymeans of numerical models that allow subsequent numerical processing to identifypatterns or measure elements identified on the surface. An even more specific area is the identification of papules on the skin, e.g. caused during an allergy test. Typically, allergy tests include a first stage in which certain positions are marked on the patient's skin where allergens that can cause an allergic reaction are placed. The markings allow these positions to be identified even when there is no allergic reaction. After the necessary time, the skin may have an allergic reaction to one or more allergens and a papule may appear. It is very important to be able to identify the presence of the papule and also to be able to dimension it. The most common ways of identifying and dimensioning papules are with very rudimentary procedures. A nurse or a physician visually checks the skin surface and at his or her discretion establishes a measurement relying at best on rulers or other visual comparison tools. These measurements are very inaccurate and depend on the person who performs the measurement so the final diagnosis is strongly influenced by subjective aspects. PCT application WO2014140215A1 describes an automated method where papules are optically scanned, identified and measured avoiding the subjectivity of the person performing the measurement. However, when the area to be scanned is large or has a high curvature, the device hasat least two scanning modules that provide data representative of portions of thesurface so that it is possible to cover the entire surface to be scanned. These separate portions of surface obtained by different modules must be joined by verifying continuity conditions. In practice this does not happen. For example, when the scanning procedure combines an illumination element such as a laser and a camera incorporating an optic with certain intrinsic parameters, lack of adjustment or optical aberrations prevent the surfaces scanned with different optics from matching.The present invention addresses this problem and provides a method which achievesthe joining of surfaces originally resulting in a discontinuity, and keeping the shape characteristics of the scanned surface intact. DESCRIPTION OF THE INVENTIONA first aspect of the present invention relates to a computer implemented method fordetecting and optionally measuring at least one protuberance, preferably one or more papules, of an area of a skin surface of a person or part of it. The method isimplemented in a computer system and comprises the steps: According to step a), the computer system receives a three-dimensional surface representation of a numerical model of an area scanned from the skin surface of a person. The skin surface of the person is scanned at least by two different scanning modules, a first scanning module and a second scanning module. Each scanning module has visual access to different portions of the skin surface of the person and may overlap. In the detailed description of the invention, an example of the invention will be described where a major region is scanned using a laser that generates an illuminated line on the surface of the person and two cameras focused on the same portion ofarea. The illuminated line generated by the laser is moved so that the illuminated linesweeps over the entire area and, the optical sensor, captures successive images so thatthe set of illuminated lines form a discretized representation of the surface.The use of two cameras focused on the same portion of the area allows greater precision in determining the position in space of each point of the line illuminated by the laser. When the surface is large or very curved, for example on an arm of a person where allergy tests are often performed, the sides of the arm are not visually accessible by one of the cameras but by the other. With the images acquired with a single camera it is possible to generate a numerical model that represents the scanned surface although it is not as accurate. The process of reconstructing the surface from the images is a known process and what it does is to place in space points or lines of the image captured by the camerathat correspond to points illuminated by the laser of the illuminated line.The data used when reconstructing the surface in the numerical model are thedirection of the laser to illuminate a given pixel in the image as well as the position andorientation of the camera capturing the image and its intrinsic data (i.e. optics, focalaxis, focal aperture and position of the optical sensor inside the camera). With these data, the point in space closest to the line determined by the laser beam and the line connecting the pixel illuminated by the beam and the focal point is found. If instead of one camera two cameras are used, then the point to be determined verifies the minimum distance between the previous two straight lines and the additional straightline generated using de image of the second camera by connecting the pixelilluminated in the image by the same laser beam of this second camera and its focal point. There are many ways to represent a surface in space. In the context of this invention isconsidered most appropriate a representation of the form ^ = ^(^, ^) where are coordinates in a given ^ domain and ^ is the variable that determines the heightover domain ^. Other forms of prepresentation of the surface such as for exampleestablishing a relation ℎ(^, ^, ^) = 0 with ℎ a scalar function will be consideredequivalent and, in these cases it is possible to establish a function ^ = ^(^, ^), even if^(·) is defined by a numerical approximation or is defined in pieces. In these cases therepresentation will be considered equivalent. Capturing a portion of the surface with the first module and capturing a differentportion of the surface with the second module, even if both surfaces have anoverlapping region, results in a discontinuity due to the lack of calibration of the first module and the second module, or any other aspect that prevents infinite precision. Itis object of the invention to succeed in modifying one of the captured surface portionsso that the surface resulting from the modification in the numerical model is continuous and, without losing the surface features captured by both modules. For this purpose it is considered by hypothesis that the discontinuity, projected on the domain ^, is approximately straight. If this hypothesis is not true, the way of solving the problem allows some curvature in the projected discontinuity since the region to be treated will be in an environment on either side of the straight line and the correction of the discontinuity can always be done even if the projection of the discontinuity is not straight since this environment is adjustable.Throughout the description, the ^ and ^ coordinates will be considered as horizontalcoordinates and, the ^ coordinate as vertical coordinate without this orientation beingessential, it is because the most usual application of scanning a surface part of aperson will be done on a base that will have a horizontal position. The base willconstitute a region corresponding to the ^ domain and, the height above the base willbe the height that determines the scanning phase on the skin surface of the patientwho places part of his body on the base. The next step is to select peaks in the computational system on both the surface captured by the first module and the surface captured by the second module. Although the surfaces have peaks due to skin irregularities, the peaks will be of larger dimensions when they correspond to the discontinuity. The selection of peaks corresponding to local maxima and minima will be on either side of the discontinuity.These peaks are used to establish a linear correlation of the form ^ = ^(^) for the(^, ^) coordinates of the selected set of peaks. That is, this step does not involve thecoordinate ^. The ^-coordinate is the longitudinal coordinate and is the coordinate along which the discontinuity is expected to lie. The ^-coordinate will be referred to as the transverse coordinate. The method according to the invention also comprises the steps:d) determining in the computer system a discretization of the longitudinal direction ^,wherein for each ^^; ^ = 1, … , ^ with ^ predetermined positive integer number,the surface representation is represented by a curve sampled in the ^ − ^ planeand, carrying out the following steps: Sampling the ^-coordinate at a plurality of points ^^allows to establish a set of points where the surface will intersect with a perpendicular plane, a cutting plane. Eachcutting plane generates a curve where ^ is the independent coordinate and ^ is thedependent coordinate. That is, the family of curves given by the plurality of points ^^will be of the form ^ = with ^^ constant value. It is considered as a practicalembodiment that the plurality of points ^^ are equidistributed.The cutting plane located at ^^ not only cuts the surface generating a curve but in the^ domain it also cuts the regression line ^ = ^(^) wherein in this case the regressionline is straight. This second cut establishes a coordinate ^^ as a point of theindependent variable on the curve ^ = ^(^^, ^), the point ^ = ^^^^, ^(^^)^ =^(^^, ^^).This point of intersection with the regression line determines two sides of the curve^ = and, it is around this point that a discontinuity search interval isestablished, the interval [^^ − ^, ^^ + ^] wherein ^ is a predetermined distance.The search is now limited to this interval of the curve. The selected portion of thecurve now defines two points, one on the left and one on the right of the interval[^^ − ^, ^^ + ^]. Between these two points a line is established so that the curve willhave sections that are over this line and other sections that are under this same line.The most distant points are selected, the most distant point over the line and the mostdistant point under the line where by hypothesis it will be considered that these twopoints are the points of discontinuity. The next step establishes a shift of the whole curve on one side of the discontinuity,not only the one selected in the interval [^^ − ^, ^^ + ^], where the shift for all ^^ isalways done on the curve section on the same side. This allows one of the surfacescaptured by the scanning modules to remain immobile and the other to be shifted toat least minimize the discontinuity.According to the invention the correction is made at least in the ^ coordinate byeliminating the difference in height. As a result of the method, now the surface of the numerical model representing the surface of at least a part of the person that has been scanned is the amended surface where at least the discontinuity has been minimized.In an embodiment, step 5) further comprises determining the distance in respect of the^ coordinate between the two points determined in step 4) and, step 6) furthermodifying the curve by shifting in the ^ direction the portion of the shifted curve, sothat, the two points are at the same location. Correcting the position of one side of the curve with an offset in the ^-coordinatereduces the vertical difference and, the additional correction in ^ makes the curvesection on one side and the curve section on the other side connect at the same point,and the additional correction in ^-coordinate makes the curve section on one side andthe curve section on the other side connect at the same point. From the data storagepoint defining the curve it is possible to remove one of the now overlapping points. According to this embodiment, the continuity of the surface generated as a joint of the modified curves is ensured. In an embodiment according to any of the previous disclosed embodiments, the three- dimensional surface representation of a numerical model further comprises a third part of the scanned area scanned at least by a third scanning module wherein steps b) to d) are carried out for shifting the curves corresponding to the side of the first part of the scanned area and for the third part of the scanned area, both shifted in respect to the second part of the scanned area, wherein the second part of the scanned area islocated between the first and the third part of the scanned area.The surface of the numerical model received after the patient surface scanning operation has three distinct surface portions, a first portion generated by the first scanner module, a second portion generated by the second scanner module, and a third portion generated by the third scanner module. In addition, there will now be a first discontinuity between the first portion and the second portion of the surface and a second discontinuity between the second portion of the surface and the third portion of the surface. In this embodiment, a displacement of one of the portions is carried out according to the described method by first establishing the correlation that determines an estimateof the projection of the first discontinuity on ^ domain and then determining thedisplacement that minimizes or corrects the discontinuity. The same applies with respect to the second discontinuity where a displacement of another of the portions is carried out according to the method described by first establishing the correlation that determines an estimate of the projection of the second discontinuity and then determining the displacement that minimizes or corrects the discontinuity. It is considered as preferred the case where the first discontinuity and the second discontinuity extend mainly along the same coordinate direction, e.g. the ^-coordinate. That is, the projection of the first discontinuity can be expressed as ^ =and the projection of the second discontinuity can be expressed as ^ = ^^(^).For the determination of the correlation ^^only peaks of the first surface portion and the second surface portion are used, and for the correlation ^^only peaks of the second surface portion and the third surface portion are used.It is also considered as preferred the embodiment in which the curve sections that are not displaced in the minimization of discontinuities or their correction are those that mainly correspond to the second surface where the second surface essentially lies,with respect to its projection in the ^ domain, between the first surface and the thirdsurface. In an embodiment according to any of the previous disclosed embodiments, three- dimensional surface representation of a numerical model of an area scanned from the skin surface of a person is by means of a combination of a laser beam generator and a camera causing the measurement of a curve located on the skin surface and being its three-dimensional surface representation of the numerical model of the scanned arealocated in the space over the domain ^.There are different technologies to determine the shape of a surface, by imageprocessing, by acoustic proximity sensors or by electromagnetic radiation with acertain wavelength, etc. All of them are able to generate data that allow to represent a surface in space. The preferred example uses a combination of a laser source and a camera as described above. The advantage of using this technique is that the laser illuminates a curve on the surface and it is this curve that is used in the image captured by the camera togenerate a curve in space. The set of captured curves define the surface consideringthat they are curved lines that correspond for example to a plane or a cut surface on the scanned surface. A surface, regardless of the mode of capture and representationcan be transformed into a surface of the form ^ = ^(^, ^) where those surfacesrepresented numerically require some kind of interpolation technique. The advantage of image capture combining a laser and a camera is that the captured curves can be oriented according to one of the coordinate directions. If for example the ^-coordinate direction is chosen then the selection of points ^^of the method can be matched to the points ^^for which a curve already exists as a result of the scanner operation. This means that the method does not require any interpolation operation and the image processing not only does not lose information but also increases the processing velocity. In an embodiment according to the previous embodiment:- the second scanning module is a combination of a laser beam generator and twoseparated cameras allowing to determine, in operative mode, the curve illuminated by the laser beam over the skin surface in the space by combining the measurement obtained by combining the information of the two images captured by the two cameras, the information about the location and orientation of the two cameras and, the information about the intrinsic parameters of the two cameras;- the first scanning module is a combination of the laser beam generator an one ofthe cameras of the second scanning module, in particular the camera having visual access to one area of the skin where the other camera has not visual access, allowing to determine, in operative mode, the curve illuminated by the laser beam over the skin surface in the space by determining a measurement using the information of the image captured by the camera, the information about the location and orientation of the camera and, the information about the intrinsic parameters of the camera. According to this embodiment, in at least one surface portion two cameras are used to determine in space each of the pixels of the curve illuminated by the laser knowing the position and orientation of the laser that has given rise to the illuminated pixels appearing in one and the other image and the intrinsic features of the two cameras. The other portion of the surface is determined only by the camera that has visualaccess to the scanned surface of the person so its accuracy is lower and generates adiscontinuity in the captured surface resulting from joining the portion captured by thetwo cameras and the portion captured by a single camera. This embodiment makes it possible to extend the surface captured by two cameras toat least one region that is only captured by one camera and the resulting surface aftercorrecting it maintains continuity.In an embodiment according to the previous embodiments using at least two cameras,wherein the third scanning module is a combination of the laser beam generator andthe camera not used in the first scanning module, in particular the camera having visual access to one area of the skin where the other camera has no visual access, allowing to determine, in operative mode, the curve illuminated by the laser beam over the skin surface in the space by determining a measurement using the information of the image captured by the camera, the information about the location and orientation of the camera and, the information about the intrinsic parameters of the camera. In the above embodiment the use of two cameras has made it possible to generate a continuous surface that not only includes the area scanned on the surface of the skin of the person with visual access of the two cameras but also the area of the skin where one of the cameras does not have visual access. In this example the scanned area is the largest possible since it takes into account three portions of scanned area, the scanned area acquired using both cameras and also the two portions of scanned area acquired by only one and the other camera where only one camera has visual access. In this case there are two discontinuities that must be compensated or corrected as described above. In an embodiment according to any of the previous embodiment, the scanning modules capture the skin surface by curves when providing a three-dimensional surface representation of a numerical model of an area scanned from the skin surface of aperson and, the discretization of the longitudinal direction ^ according to step d)corresponds to the set of curves originally provided by the three-dimensional surface representation of the numerical model of an area of the skin surface.In this embodiment, the example already disclosed is explicitly considered where theline illuminated by the laser generates a curve in space which in projection on Ωdomain is made to coincide with one of the coordinates, for example the ^-coordinate.This allows the captured data to already admit a representation of the form ^ =^(^, ^) so that for the ^^ where curves have been captured, the curve used in thecompensation or correction of the discontinuity has a representation of the form ^ =^(^^, ^) so that no interpolation operation is necessary. That is, the stored raw dataare those that can be directly processed in the stages of the method that determine the displacement that reduces or corrects the discontinuity. This example embodiment has been found to be particularly fast in image processing, so that it is especially suitable for those devices that must provide a result at the time, when the person or patient is still in the test that requires the capture of the surface.Any of the described examples is of interest in obtaining a representative surface ofthe skin surface of a patient on whom an allergen test is being performed. Following the application of allergens to various points on the skin, capturing the surface by reproducing surface features such as the presence of papules allows identification of whether papules are present and how large they are. That is, it is possible to identify which allergens generate a reaction on the patient's skin and to quantify the reactionin an automated way.A second aspect of the invention is a device adapted to scan at least part of the skinsurface of a person and, a computer system, the computer system adapted to carryout any of the previous disclosed methods. A third aspect of the invention is a computer program product comprising instructionswhich, when the program is executed by a computer system, cause the computer tocarry out the steps of any of the previous disclosed methods.DESCRIPTION OF THE DRAWINGS These and other features and advantages of the invention will be seen more clearly from the following detailed description of a preferred embodiment provided only by way of illustrative and non-limiting example in reference to the attached drawings.Figure 1 This figure shows an embodiment of the invention formed by a deviceintended for capturing the shape of the surface of a part of a user's body, namely the arm. The practical application of the device shown inthis figure is the capture of the surface of the skin of a user who has undergone an allergy test. Those parts that react to an allergen generate a papule that is automatically measured by the device from the shape of the captured surface.Figure 2 This figure shows schematically the device of the previous figure, in anelevation view. This schematic figure shows two modules for capturing the shape of the patient's skin surface located on one and the other side of the arm, also schematically depicted.Figure 3 This figure shows a perspective representation of the numerical modelrepresenting the surface of the arm as captured. In this example, three areas are shown with a discontinuity extending mainly along the longitudinal direction X since each area has been captured with a different capture module causing a discrepancy where the surfacesshould be continuous.Figure 4 This figure shows a cloud of points corresponding to local maxima andminima of the captured surfaces located near one of the discontinuities. The point cloud is used to establish a linear correlation identified as ^ = ^(^).Figure 5 This figure shows a detail of the region delimited around a pointestablished by the line correlation and where it is studied what displacement is finally applied to the curves in the ^ − ^ variables thatup to now do not verify continuity.Figure 6 This figure is like figure 3 but now the surface of the numerical modelis the result of having eliminated the discontinuity according to the described method. DETAILED DESCRIPTION OF THE INVENTION As will be appreciated by one skilled in the art, aspects of the present invention may be embodied as a system, method or computer program product. Figure 1 and figure 2 shows an embodiment of a device comprising a computersystem (S) adapted to carry out the disclosed method for detecting and optionallymeasuring at least one protuberance of an area of a skin surface (SS) of a person orpart of it. The device also comprises a first scanning module (1) and a second scanningmodule (2), the two scanning modules (1, 2) adapted to capture the shape of an areaof a skin surface of a person or part of it. In particular, the area of the arm of a person.Figure 2 also shows a third camera (3) which in this case is an RGB camera in a zenithalposition and which allows skin surface color (SS) to be acquired.Figure 2 also shows a computing system (S) connected to the hardware elements suchas the cameras or the laser emitter. This computational system (S) allows the laserbeam to be synchronized with the images captured by the cameras. Similarly, once theimages have been captured by the cameras, the computational system (S) is adaptedto process the images in order to generate the surfaces that form part of the numerical model representing the patient's skin surface.The same computing system (S), according to this embodiment, is the one that carriesout the described method, eliminating the discontinuities due to the acquisition of the whole surface with different scanning modules.Although the computational system (S) is shown separate from the arc-shaped device,this representation is schematic and this particular embodiment incorporates all the elements of the computational system integrated into the arc-shaped casing.In this embodiment, the device is configured in the form of an arc and is positioned ona base (B), the base (B) being used by the person to rest the arm on which the captureof its surface is to be carried out. In one example, the arm is that of a person to whom various allergens have previously been applied at predetermined points on the arm to induce an allergic reaction, if any, resulting in papules in the form of bumps emerging from the surface of the skin. The arc-shaped device comprises two obliquely positioned cameras, one on each side of the arm as shown in Figure 2, so that each camera has visual access to a different area of the arm. The area accessed by the left camera partially overlaps with the area accessed by the right camera. Thus, each camera has visual access to an area that can be divided into two areas, one to which only one camera has visual access and one to which both cameras have visual access. As a result, it is possible to capture three different regions, a left region that can only be captured by the left camera, a region centered along the transverse Y direction that can be captured by both cameras, and a third region that can only be captured by the right camera. The central area, which can be captured by both cameras, is captured in this way, using both cameras as this is the most accurate. As described above, the skin surface shape scanning modules combine a camera and a laser emitter designed to illuminate a line on the surface. The line captured by the camera makes it possible to determine, for each pixel, the position in space of the point on the skin that illuminated the camera pixel. This determination is made by knowing the direction of the laser in space and the intrinsic and extrinsic characteristics of the camera, such as the focal length, the position of the focal point in space and the position of the sensor with respect to the focal point. The direction connecting the illuminated pixel and the focal point results in a line that intersects the plane in which the laser is moving, resulting in the line illuminated by the laser. When two cameras are combined, two lines in space are determined and the point that verifies the minimum distance between them is taken as the resulting point. This is the combination that provides a more accurate position of a point on the surface in space. While the use of two cameras is described, the use of three scanning modules is described. Thus, the combination of the laser emitter with a camera to capture thearea that only the first camera has visual access to is called the first scanningmodule (1). The combination of two cameras and the laser emitter results in adifferent combination of hardware and processing mode and is therefore referred toas the second scanning module (2). Finally, the combination of the second camera withthe laser emitter in determining the shape of the surface, where only the secondcamera has visual access, is identified as the third scanning module (3). That is, thethree scanning modules (1, 2, 3) may share hardware elements, but they are distinctbecause the combination is different and also the way in which the captured information is processed is different. Figure 1 also shows the three spatial directions as they will be described. The orientation of the reference directions may change, but if the embodiments determine the same points, relationships and ultimately correction offsets, then it will be understood that such actions are equivalent. The same applies to mathematical expressions. A given function can be expressed in different ways, but if the relationship it establishes between its variables is the same, then the equations will also be considered equivalent.In a first step, according to this particular embodiment, the three scanning modules (1,2, 3) scan three areas of the arm.Because each of the scanning modules has a different accuracy and a different way of combining the acquisition hardware, the result is three surface sections that do not verify continuity.Figure 3 shows a three-dimensional surface representation (SR) received by thecomputational system (S). In this three-dimensional surface representation (SR) of anumerical model it is observed that there are two discontinuities or jumps whichextend in a direction almost parallel to the longitudinal X direction. Ω is the domain in^ and ^ coordinates which at least corresponds to the projection of the surface on thehorizontal plane X-Y. The numerical model contains the information that allows the points of the recorded surface to be stored, represented or manipulated. The numerical model includes at least the coordinates where the pixels illuminated by the laser have made it possible to define a point in space as a point on the skin surface. In this Figure 3, the X direction is identified as the longitudinal direction and is the direction in which the person has approximately oriented the arm. The transverse direction Y is the direction perpendicular to the longitudinal direction, where the longitudinal direction X and the transverse direction Y are considered to be horizontal. The vertical Z direction is the direction of the action of gravity, taken as a reference to the vertical or horizontal direction. The following describes how to eliminate the discontinuity between the first part ofthe scanned area (A1) scanned by the first scanning module (1) and the second part ofthe scanned area (A2) scanned by the second scanning module (2).In this case, the second part of the scanned area (A2) is the area that is taken as areference, so that parts of the first part of the scanned area (A1) are moved toeliminate the discontinuity as it will be disclosed.The same procedure between the second part of the scanned area (A2) and the thirdpart of the scanned area (A3) will eliminate the other longitudinal discontinuity. In thiscase, it is also useful to leave the central area unchanged and to impose thedisplacement on the third part of the scanned area (A3), since it is the central area thathas been scanned by two cameras and therefore its shape and position in space is assumed to be more precise.The three-dimensional surface representation (SR) of the numerical model isconsidered as a function of the form ^ = ^(^, ^), i.e. for each coordinate of thedomain (^, ^), the ^-coordinate is determined which positions the height at which thesurface is located. According to step b), a plurality of ^-coordinates are selected determining curves inthe ^ − ^ plane. For each curve, peaks (p1, p2) are determined wherein such peak maybe maximum or minimum values. In this process it does not matter whether the valuesare maximum or minimum, it only matters that it is an extreme given that it identifies positions where there is a discontinuity. These extremes can be on either side of the discontinuity.From the cloud of extreme points, only the ^ and ^ coordinates are taken, the ^coordinate corresponding to the position of the ^ − ^ plane in which the section wasmade and the ^ coordinate corresponding to the transverse location of the extremepoint.The point cloud makes it possible to establish a correlation of the form ^ = ^(^),preferably linear, so that the correlation establishes the line along which the discontinuity is expected to be found. That is, the discontinuity is expected to be close to parallel to the X direction, but it is not parallel. This makes it possible to establish,for each ^^; ^ = 1, … , ^ with ^ predetermined positive integer number, of adiscretization of the longitudinal coordinate X, in a section with a ^ − ^ plane with thesurfaces, two curves, one from each region to be joined removing the discontinuity,where in particular the extreme points to be joined will be. Figure 4 shows an exampleof a linear correlation determining ^ = ^(^). If the cloud of points shows a curvedtendency, other expressions of ^(^) can be used to give a better fit. To perform the joint method between surfaces with discontinuity, a partition of the longitudinal direction X is generated. If the data structure storing the numerical modelalready contains a distribution of lines or points of measurement in ^ and ^, it is mostefficient if the partition corresponds to the points where the values already exist. This allows a much more efficient method as it avoids, for example, having to use interpolation techniques to operate on points where the data structure does not contain values.For each point ^^ of the longitudinal discretization, the curve ^ = ^(^^, ^) isestablished and can be represented as the curve resulting from the intersection of aplane parallel to the ^ − ^ plane with the surfaces.The search for the relevant points for performing the joint is limited to an interval inthe vicinity of the discontinuity. The discontinuity is located at coordinates (^^, ^^) =^^^, ^(^^)^, so once a distance ^ has been defined, the search interval will be(^^ − ^, ^^ + ^) on the curve ^ = extending in ^ − ^ coordinates. Forinstance, in the expression ^ = ^^ is a fixed value and therefore ^ onlydepends on ^.The vicinity of the discontinuity has two ends at ^^ − ^ and ^^ + ^. The next step is todetermine the line connecting the coordinates ^^^ − ^, ^(^^, ^^ − ^)^ and^^^ + ^, ^(^^, ^^ + ^)^. Figure 5 shows two portions of the curve, a first portion of thecurve corresponding to the first area (A1) and a second portion of the curvecorresponding to the second area (A2), both portions connected in the discontinuitythat, in this case, is located at ^^ = ^(^^) and, of course, withing the interval(^^ − ^, ^^ + ^). The line connecting the two end points is shown by an obliquedashed line. Above this dashed line, p1 is the point with the greatest distance above the dashed line and, correspondingly, p2 is the point with the greatest distance below the dashed line. Auxiliary dashed lines, perpendicular to the dashed line connectingthe two end points, show the indicated distances. The expressions “on” “over” or“above” are in respect to the z coordinate in the increasing direction and,correspondingly, “under” or “below” are in respect to the ^ coordinate but in thedecreasing direction. The determination of the extreme points p1 and p2 makes it possible to determine thejump of the discontinuity, in particular the separation in the Y direction (identified as^) and the height distance ^ of the separation in the Z direction. It has been shownthat the main separation is in the ^ direction, so if this separation is corrected, themain cause of the discontinuity disappears. In this embodiment, the curve on the left is moved to the right by the y distance andraised by the ^ distance so that the curve on the left is connected to the curve on theright.Figure 6 shows the three-dimensional surface representation (SR) of a numericalmodel once the first part of the scanned area (A1) has been connected to the secondpart of the scanned area (A2) and, also the third part of the scanned area (A3) hasbeen connected to the second part of the scanned area (A2) removing the twodiscontinuities. A comparison of Figure 3 and Figure 6 shows that two papules are represented in both numerical models, before and after the removal of the two discontinuities, indicating that the method does not alter the shape of the papules even when the surface is treated cut by cut. The method described does not require surfaces to be captured by a combination of a laser beam and one or more cameras; the scanning modules can use other physical principles such as proximity sensors, radar techniques, etc. The method allows surfaces scanned by two different scanning modules providing two mismatched surfaces to be matched, resulting in a new surface that is free of discontinuities and retains the surface structure. Of particular interest is the ability to represent the surface of the skin, including the papules produced in an allergy test. Papules are lumpy areas that are usually measured manually or roughly. Having a numerical model of the surface representation allows papules to be identified automatically and to be sized in a non-subjective way.Techniques to identify papules are segmentation or comparison between the surfacewith papules and a surface that, after a filtering operation, eliminates the papules. The difference reveals the structures of each papule for subsequent sizing.

Claims

CLAIMS1.- Computer implemented method for detecting and optionally measuring at leastone protuberance of an area of a skin surface (SS) of a person or part of it, by means ofa computer system (S); the method comprising the steps:a) receive in the computer system (S) at least a three-dimensional surfacerepresentation (SR) of a numerical model of an area scanned from the skinsurface (SS) of a person, a first part of the scanned area (A1) scanned at least by afirst scanning module (1) and a second part of the scanned area (A2) scanned atleast by a second scanning module (2) and, wherein the three-dimensional surfacerepresentation (SR) may be expressed by means of two-dimensional coordinates(^, ^), a longitudinal direction (^) and a transversal direction (^), in apredetermined domain Ω and, and a function using a third-dimensional coordinate^ = ^(^, ^) representing the height of the surface representation (SR) of the skinsurface (SS) as a function of the two-dimensional coordinates (^, ^);b) selecting peaks (p1, p2) in the computer system (S), local maximum values of ^coordinate and local minimum values of ^ coordinate, along the transversaldirection (^) for a plurality of longitudinal coordinates (^);c) determining in the computer system (S), a linear correlation ^ = ^(^) from the(^, ^) coordinates of the peaks (p1, p2) selected in the previous step;d) determining in the computer (S) system a discretization of the longitudinaldirection (^), wherein for each ^^; ^ = 1, … , ^ with ^ predetermined positiveinteger number, the surface representation (SR) is represented by a curve (c)sampled in the ^ − ^ plane and, carrying out the following steps:1) representing the surface representation (SR) by a curve (c) ^ = ^(^^, ^);2) determining an interval [^^ − ^, ^^ + ^] wherein ^^ is determined by thecorrelation ^^ = ^(^^), preferably a linear correlation, and ^ is apredetermined distance; 3) determining a line (l) in the ^ − ^ plane between the two end points of theinterval, i.e., (^^ − ^, ^(^^, ^^ − ^)) and (^^ + ^, ^(^^, ^^ + ^));4) determining a first point (p1) of the curve (c) with the highest distance overthe line (l) and a second point (p2) with the highest distance under theline (l), the over and under property in respect of the ^ coordinate;5) determining the distance in respect of the ^ coordinate between the twopoints (p1, p2) of the previous step;6) modifying the curve (c) by shifting in the ^ direction the portion of thecurve (c) located at one predetermined side, the same side for all ^^;e) providing a new three-dimensional surface representation (SR) of the numericalmodel, the surface representation (SR) defined by means of the modified set ofcurves (c).2.- A method according to claim 1, wherein step 5) further comprises determining thedistance in respect of the ^ coordinate between the two points (p1, p2) determined instep 4) and, step 6) further modifying the curve (c) by shifting in the ^ direction theportion of the shifted curve, so that, the two points (p1, p2) are at the same location.3.- A method according to claim 1 or 2, wherein the three-dimensional surfacerepresentation (SR) of a numerical model further comprises a third part of the scannedarea (A3) scanned at least by a third scanning module (3) wherein steps b) to d) arecarried out for shifting the curves (c) corresponding to the side of the first part of thescanned area (A1) and for the third part of the scanned area (A3), both shifted inrespect to the second part of the scanned area (A2), wherein the second part of thescanned area (A2) is located between the first (A1) and the third (A3) part of thescanned area.4.- A method according to any of previous claims, wherein three-dimensional surfacerepresentation (SR) of a numerical model of an area scanned from the skin surface (SS)of a person is by means of a combination of a laser beam generator (4) and acamera (1, 2, 3) causing the measurement of a curve located on the skin surface (SS)and being its three-dimensional surface representation (SR) of the numerical model ofthe scanned area located in the space over the domain Ω.5.- A method according to the previous claim, wherein:- the second scanning module (2) is a combination of a laser beam generator (B) andtwo separated cameras (2.1, 2.2) allowing to determine, in operative mode, thecurve (c) illuminated by the laser beam over the skin surface (SS) in the space bycombining the measurement obtained by combining the information of the two images captured by the two cameras (2.1, 2.2), the information about the locationand orientation of the two cameras (2.1, 2.2) and, the information about theintrinsic parameters of the two cameras (2.1, 2.2);- the first scanning module (1) is a combination of the laser beam generator (B) anone of the cameras (2.1) of the second scanning module (2), in particular thecamera (2.1) having visual access to one area of the skin where the othercamera (2.2) has not visual access, allowing to determine, in operative mode, thecurve (c) illuminated by the laser beam over the skin surface (SS) in the space bydetermining a measurement using the information of the image captured by the camera (2.1), the information about the location and orientation of thecamera (2.1) and, the information about the intrinsic parameters of thecamera (2.1).6.- A method according to claims 4 and 5, wherein- -the third scanning module (3) is a combination of the laser beam generator (B)and the camera (2.2) not used in the first scanning module (1), in particular thecamera (2.2) having visual access to one area of the skin where the othercamera (2.1) has no visual access, allowing to determine, in operative mode, thecurve (c) illuminated by the laser beam over the skin surface (SS) in the space bydetermining a measurement using the information of the image captured by the camera (2.2), the information about the location and orientation of thecamera (2.2) and, the information about the intrinsic parameters of thecamera (2.2).7.- A method according to any of the previous claims, wherein the scanningmodules (1, 2, 3) capture the skin surface (SS) by curves (c) when providing a three-dimensional surface representation (SR) of a numerical model of an area scanned fromthe skin surface (SS) of a person and, the discretization of the longitudinal direction ^according to step d) corresponds to the set of curves (c) originally provided by thethree-dimensional surface representation (SR) of the numerical model of an area ofthe skin surface (SS).

8. A device comprising at least a first scanning module (1) and a second scanningmodule (2) adapted to scan at least part of the skin surface (SS) of a person and, acomputer system (S), the computer system adapted to carry out the method accordingto any of previous claims.9.- A device according to the previous claim, the device further comprising a thirdscanning module (3) wherein the computer system (S) is adapted to carry out amethod according to any of claims 3 to 7.10.- A computer program product comprising instructions which, when the program isexecuted by a computer system (S), cause the computer to carry out the steps of themethod of any of claims 1 to 7.

Images