WO2019179809A1

WO2019179809A1 - Method for classifying a vehicle seat and classification system for classifying a vehicle seat

Info

Publication number: WO2019179809A1
Application number: PCT/EP2019/056021
Authority: WO
Inventors: Matthias Franz; Thomas Holzinger
Original assignee: Bayerische Motoren Werke Aktiengesellschaft
Priority date: 2018-03-23
Filing date: 2019-03-11
Publication date: 2019-09-26
Also published as: DE102018204495A1

Abstract

Exemplary embodiments of the invention relate to a method for classifying a vehicle seat in a motor vehicle. The method comprises positioning at least one standard weight on the vehicle seat and creating at least one camera image of the vehicle seat with the standard weight. Superposing model data of the vehicle seat onto the camera image makes it possible to classify the vehicle seat by evaluating the camera image onto which model data have been superposed. Further exemplary embodiments relate to a classification system that enables the classification of a vehicle seat.

Description

A method of classifying a vehicle seat and a classification system for classifying a vehicle seat

description

Embodiments of the present disclosure relate to a method of classifying vehicle seats and a classification system configured to classify vehicle seats.

A vehicle seat, for example, a seat of a motor vehicle, a motor vehicle, a rail nenfahrzeugs, an aircraft or a watercraft. In the manufacture of such vehicles, e.g. Motor vehicles, inter alia, a quality analysis of the vehicle seats installed in the motor vehicle is provided. For example, workstations such as seams or material quality can be tested. It is also possible to test a firmness of the seat cushion, which is approximately relevant for the comfort of a user of the vehicle seat.

Seat analyzes are carried out predominantly subjectively. Test persons can assess the sitting feeling on a vehicle seat and indicate the seat comfort they perceive. In such analyzes with volunteers, measuring devices such as pressure measuring mats can be included in order to objectify the results. Based on the result of a seat comfort analysis, a vehicle seat may be classified, for example, it may be determined whether it meets predetermined comfort requirements or not.

From known seat comfort analyzes, however, it may emerge that an analysis result is not reproducible due to a subjectivity of the evaluations by the subjects, for example. In particular, an objective classification of seating comfort classes can only be limited. The requirement of test subjects present can restrict the performance of seat comfort analyzes in time and be time consuming and can be costly.

In particular, a comparison between several seats and a differentiated classification in fine gradations is difficult for subjects to perform. For a series monitoring in a production of vehicles in a production plant, a seat comfort analysis with subjects may not be possible due to a large number of manufactured vehicles in series production. There is a need to provide an improved concept for classifying vehicle seats. This requirement is taken into account by the method and the system according to the independent claims.

Embodiments relate to a method for classifying a vehicle seat of a motor vehicle. The method comprises positioning at least one standard weight on the vehicle seat and creating at least one camera image of the vehicle seat with the standard weight. Thus, a camera image is created that shows at least a portion of the vehicle seat on which the standard weight is positioned. Furthermore, the method comprises superposing the camera image with model data of the vehicle seat. The model data has classification ranges. A classification of the vehicle seat takes place by means of evaluation of the camera image superimposed with the model data.

The standard weight can deform the vehicle seat and this deformation can be determined on the camera image. The model data may include, for example, data on various degrees of deformation. On the basis of the evaluation of the deformation of the vehicle seat with the model data, such as by comparing the actual deformation with the data at various degrees of deformation, the vehicle seat can be classified.

It is possible that the model data includes a mark, which is displayed on a screen, for example, on which a live image of the vehicle seat is displayed (for example, the vehicle seat is filmed and model data superimposed on the filmed image). By marking, it is possible to position / set the standard weight at a predetermined location, and it is possible to reproduce the positioning without fluctuations due to changing positions of the standard weight. Thus, a guidance for positioning can be provided. For example, a light contour can be projected to the predetermined location.

Furthermore, it is possible to create an image of the vehicle seat without the standard weight before positioning. The image without standard weight can eg be used to create model data or even display the model data. It is possible to use the no-weight image for comparison with the standard weight image (on which the vehicle seat is deformed) and to compare how much the vehicle seat or a seat cushion deforms. For example, a distance between a deformed and an undeformed seat surface can be determined. For example, a difference (delta) between a first position of the seat surface without standard weight and a second position of the seat surface with standard weight (eg: delta = | p_l_with_normal_weight-P_2_: with_nonweight |). By the picture each with and without standard weight a delta can be determined with the same seat with the help of the picture. The classification of the vehicle seat may include that a strength of a seat cushion is classi fied, for example, the strength of a seat. A seat cushion may be a padding of the vehicle seat on a seat, on a backrest, on a headrest. In Fahrzeugit zen strength of a seat can be designed so that it is as comfortable as possible for a user of the driving convincing seat and in this case is between a too high hardness and too low hardness. It can be provided a standard strength of the seat, for example, with a me chanical tension of 6 kPa, for example, with a standard tolerance of 20%. A strength may depend on a foam part, a reference voltage or a suspension of the vehicle seat. Due to tolerances in the production process, the strengths may be too soft or too hard for individual vehicle seats, for example, so that they deviate from the standard strength.

The classification according to the proposed method makes it possible to classify vehicle seats efficiently and inexpensively, for example with regard to the strength of seat cushions. Thus, it is possible to recognize vehicle seats that deviate from a given standard, and for example not to be installed in motor vehicles, or rule out of their respective motor vehicles. The proposed method makes it possible that a classification of the vehicle seat takes place directly in a motor vehicle, the vehicle seat does not have to be removed from the vehicle to do so. A classification of vehicle seats according to the method is carried out by means of digitalization in vehicle production or production, in particular using augmented reality, virtual data being added to a camera image.

The classification can be carried out with respect to different quality classes or comfort classes. For example, vehicle seats that are subject to a desired specification of strength, e.g. the standard strength of their seat upholstery by less than 5% or less than 3% are assigned to a highest quality class, vehicle seats deviating from a target specification by more than 5% and less than 10% are assigned to a middle quality class and vehicle seats, which deviate from a target specification by more than 10% to a low quality class.

The classification of vehicle seats is carried out according to the method using a standard weight and a camera image. By the standard weight, which is positioned on the vehicle seat, the vehicle seat deforms at the respective location depending on, for example, the strength of his seat cushion. Since the mass of the standard weight is known, can be deduced based on the strength or the degree of deformation on a strength of the seat cushion. In order to determine the deformation caused by the standard weight, a camera image of the Vehicle seat with the standard weight created and the camera image by model data, such as virtual elements, for example, based on augmented reality methods extended.

Vehicle seat model data may be design data or digital models of the vehicle seat. Model data may be, for example, computer aided design (CAD) data. The Mo del ldaten can represent shapes, contours and size of the vehicle seat, if this is not burdened with a standard weight, especially if it is unloaded. The model data can represent contours of the vehicle seat, for example, which, in the event that the vehicle seat to be classified is unloaded, ie no weight is positioned on the vehicle seat, trace the contours of the vehicle seat to be classified. The model data may include markings indicative of a deformation of the vehicle seat by the standard weight at a standard strength of the seat cushion. For example, if the actual deformation deviates from the mark, it may be determined that the seat cushion of the vehicle seat has a different strength than the standard and the vehicle seat is classified accordingly.

The overlaying of the model data on the camera image of the vehicle seat to be classified may include adapting the model data in relation to a size, a rotation and / or a translation as a function of a camera perspective from which he takes place the camera image. The vehicle seat to be classified may be referred to as a real vehicle seat, whereas the model data may be referred to as a virtual vehicle seat. For example, overlaying may include maximizing a correlation function of the model data and the real vehicle seat. The real vehicle seat and its position on the camera image can be detected by means of tracking software or image recognition. For example, a further camera image of the unloaded real vehicle seat can be created before the positioning of the standard weight for the superposition with model data. As a result, it may be particularly easy to superimpose the model data on the camera image.

The model data of the vehicle seat, which are superimposed on the camera image of the vehicle seat with the standard weight, can provide a classification basis. For example, it can be evaluated for classifying by what degree, for example, a contour or a surface of the real vehicle seat deviates from a contour or surface of the model data of the vehicle seat. A deviation results from the deformation of a seat cushion by the standard weight. It is possible that the model data have a standard deviation, for example a standard deviation range, for a load of the vehicle seat with the standard weight. If it is detected on the camera image extended by the model data that the deviation is within the standard deviation, the vehicle seat can be classified as comfortable become. If the determined deviation is outside the standard deviation, the driving seat can be classified, for example, as uncomfortable.

The classification of the vehicle seat can be determined manually on the basis of the image detected on the camera image deviation of a shape of the vehicle seat of superimposed model data. It is also possible that a classification takes place automatically, for example by means of a pattern recognition software, which is designed to determine the degree of deviation of the vehicle seat from the model data. For example, in an automated classification, if a deviation is within the standard deviation, a green signal may be output so that a quality of the vehicle seat is OK or a red signal is output if a deviation is outside the standard deviation , A yellow or orange signal may be output if a deviation is in a limit adjacent to the standard deviation.

The proposed method makes it possible to classify vehicle seats by means of a camera evaluation. As a result, a quality or a user comfort of the vehicle seat can be classified, which may depend on a strength of a seat cushion. Such Klassifi coding is possible in particular in an already manufactured vehicle, without the vehicle seat, for example, must be changed or removed. A classification is also possible from a distance in which there is an optical connection between the seat and camera, example, through a window of a climate chamber, which are reduced by the method technical requirements of the camera, since it can be operated outside the climate chamber. The method makes classification quick and efficient, therefore it can be used in a series production for vehicle seat quality assurance. Insbesonde re in an automated classification, for example, to carry out the method only a manual positioning of the standard weight on the vehicle seat necessary. As a result, steps in a classification can be reduced, in particular a classification can be carried out without test subjects and on the basis of an objective evaluation scale.

According to one embodiment of the method, the standard weight is positioned on a seat of the driving convincing seat. The seat can have approximately a foam part, which can be coated with a coating of a textile material or of a leather material. For example, the seat can be set for classification at a predetermined inclination angle. The predetermined angle of inclination may, for example, correspond to a co-location (construction position) of the vehicle seat, or be inclined by 0 °, 10 °, or 20 ° with respect to a horizontal floor surface. The standard weight can also be positioned on all other seat cushion surfaces of the vehicle seat. For example, it can be positioned on a backrest or a headrest. According to the method, camera images of the backrest or the head restraint can in each case be created in these cases and the model data comprise these parts of the vehicle seat. For example, a backrest can be placed vertically, so that a positioned standard weight remains in the previously provided position and does not slip off the backrest. It is also possible to attach the standard weight for classifying a strength of the backrest by means of a suitable position onshalterung on the headrest. To classify a strength of the headrest, a standard weight can be positioned analogously to the headrest. An angle of inclination of the backrest can also correspond to a co-location, or be inclined by 10 °, 20 °, or 30 ° relative to a vertical. For example, an adjustment of movable components of the vehicle seat may correspond to a representation of the model data, so that an overlay is possible. By the method so vehicle seats with respect to their various components, for example, all upholstery can be classified. This makes a more differentiated objective classification possible.

According to one embodiment of the method for the superimposition of the camera image of the driving convincing seat assigned model data is used. The model data has at least information about the unloaded vehicle seat. Depending on the construction and materials used for the seat surface, the model data may have different classification ranges. For example, it may be provided in a vehicle seat for a sports car, that a strength of the Fahrzeugit ze is higher than, for example, in a vehicle seat for a minibus. By means of image recognition software, it can be recognized which vehicle seat model is involved and, depending on the detected vehicle seat, model data assigned to the vehicle seat can be loaded and used from a database, for example. The information about the unloaded vehicle seat are about a shape contour or surface of the vehicle seat when no weight loaded the driving convincing seat. For example, the model data correspond to a shape of the vehicle seat to be classified if it is not loaded with the standard weight.

According to a development of the method, the classification areas are assigned to a respective standard weight and detect the degree of plastic deformation of a surface of the vehicle seat by the respective standard weight. For example, an identification mark, for example a QR code, can be attached to a standard weight used, by means of which the respective standard weight can be recognized by means of image recognition software. For example, due to the recognized standard weight, associated classification ranges can be used. For example, classification ranges may be selected differently for a first standard weight than for a second standard weight if the first standard weight is a lower mass, for example has as the second standard weight. A recognition of the standard weight can be automated by the identifi cation mark on the standard weight about easier.

According to a further development of the method, at least three classification ranges are used, wherein a strength greater than the standard strength is assigned to a first classification range, a second classification range is assigned a strength corresponding to the standard strength, and a third classification range has a lower strength than a standard strength of a vehicle seat surface the standard strength is assigned. For example, a deformation of the vehicle seat by the standard weight may cause a standard deviation of a seat surface to model data of the vehicle seat. In this case, the classification may associate the vehicle seat with the second classification area, so that it may be known that the vehicle seat provides high user comfort. If a deformation is outside the standard deviation, the classification may assign the vehicle seat to the first or third classification area, so that it may be known that a user comfort of the vehicle seat may be limited.

According to a further development, the method further comprises determining a measurement temperature and a measurement humidity when creating the camera image and normalizing a classification of the vehicle seat to a standard temperature and a standard humidity using a material table comprising at least one material of the vehicle seat. For example, different materials at different temperature or humidity at a load by the standard weight depending on the respective temperature and humidity under different deform. For example, information about how a strength of a material of a seat cushion changes depending on the temperature can be provided in a material table. For example, if a measurement is made for a classification at a first temperature other than a standard temperature, the result of the classification based on the material table may be converted or adjusted or normalized to a result that would have been achieved at a standard temperature. In this case, a relationship of the classification result and the temperature dependence of a material hardness of a cushioning material may be approximately non-linear, since other factors, e.g. a reference voltage of the vehicle seat affect the classification result. A temperature may be determined, for example, by means of an optical thermometer, such as a laser thermometer. By taking into account the temperature and / or the humidity, it is possible to enable an increased objectivity of a classification of vehicle seats also in different measuring environments.

According to a development of the method, it is provided that the standard weight has at least one marking which, for comparison with the model data and assignment of a classification area is formed. A marker can have multiple marker lines. A jewei time marking line can be assigned to a respective strength of a seat cushion. If the standard weight sinks into the seat cushion during positioning on the seat cushion, for example, one of the respective marking lines may lie on a contour line of the model data of the vehicle seat superimposed on the camera image. Due to the development, it is possible to determine a strength of the seat cushion thereby, the one marking line, which lies on the contour line of the model data, read. As a result, a classification can be particularly easily possible, for example, a classification can be carried out particularly easily by a trained computer program with image recognition. For example, the different marking lines have different colors. By way of example, the marking has a color spectrum which changes continuously over a spatial extent, so that a classification of the vehicle seat with continuous classification values is possible.

According to a development of the method, the camera image is created after a period of time longer than, for example, 60 s after the positioning of the standard weight on the vehicle seat. The duration may be longer than 10 seconds, longer than 30 seconds, longer than 60 seconds, longer than 90 seconds, or longer than 120 seconds. It may be that a seat pad of the vehicle seat deforms only after a certain time, so that a static deformation sets only after a predetermined period _· Waiting for the creation of the camera image by the time until the static deformation has set, can Quality of a classification of the vehicle seat can be increased. The time can be selected depending on a material of the vehicle seat so that the camera image is only created when a stati cal deformation of the material has been adjusted.

According to a development of the method, the standard weight has a mass of at least 3 kg and a maximum of 20 kg. For example, the standard weight has a mass of more than one kilo, more than two kilos, more than five kilos or more than ten kilos. For example, the standard weight has a mass of less than 18 kilos, less than 15 kilos or less than ten kilos. For example, the standard weight has a predetermined size. A size of a Aufla gefläche the standard weight on the vehicle seat may correspond to a size of a seat bone, for example, have a diameter of more than 5 cm and less than 15 cm. This can improve a realistic classification. For example, a height of the standard weight may be used as a mark for evaluation with the model data, a standard deviation of a deformation may be defined by an upper edge of the standard weight in a vehicle seat having a standard strength on a respective shape contour of the model data. According to one embodiment of the method, the standard weight is positioned on a headrest, a backrest and / or a side support cushion of the vehicle seat. For example, the standard weight depending on a distance of the side support cushion of the vehicle seat dimensi oned, so that it rests on the two Seitenhaltpolstem the vehicle seat. In this way, Kings also strengths or comfort classes of side cushion, headrest and / or backrest of the vehicle seat can be classified.

According to one development of the method, two standard weights for classifying a vehicle seat are positioned at two different locations of the vehicle seat. For example, a plurality of weights may also be positi on a plurality of locations of the vehicle seat. The standard weight may consist of a plurality of elements that can emulate a body of a user of the vehicle seat. Such a standard weight may partly have the form of a dummy. By continuing the strength of various components of the vehicle seat can be classified about the same time. For example, various standard weights of various masses can be positioned on the seat surface of the vehicle seat, so that deformation of the seat surface with unequal mass distribution can be examined and the vehicle seat can be classified accordingly. In this way, the behavior of the vehicle seat in cornering with centrifugal forces can be classified.

According to a further development of the method, the classification further comprises a check of a functionality of a movement device of the vehicle seat, wherein a movement of the standard weight caused by the activated movement device is determined by means of evaluation of at least two camera images, the functionality being classified based on the movement of the standard weight. Vehicle seats may include active movement devices, such as massage functions. For example, the seat surface of a vehicle seat may include pneumatic chambers that can move a pelvic area of a user of the vehicle seat. A classification of the moving means may be to test a functionality of the moving means. When the moving means is activated and the standard weight is positioned in a certain vicinity, for example above the moving means, the standard weight is moved following a movement of the moving means. The model data may have an upper limit and a lower limit within which the standard weight must at least move with an activated mover in order for the mover to be classified as functional. The camera images can then be determined if the standard weight is at extreme positions of the movement, for example at extreme positions of a flub amplitude. For example, a plurality of camera images can be determined to facilitate the detection of the extreme positions. For example, a video recording with superimposed or overlaid model data can be used become. Due to the development, it is possible to classify other functions of the vehicle seat in addition to a strength of a seat cushion of a vehicle seat.

Another aspect of the disclosure relates to a classification system for classifying a vehicle seat. The classification system comprises at least a standard weight and a Bildaus evaluation device. The image evaluation device comprises at least one camera and an evaluation device, wherein the evaluation device has model data for at least one driving seat and is adapted to superimpose a camera image of the camera with the model data. The classification system is thus designed to export a method described above. A classification system may be provided in a portable device, such as a tablet, or may be a distributed or a stationary system. For example, the camera may be connected to the evaluation device via a wireless connection, such as WLAN or Bluetooth, or may be connected via a network connection. As a result, about one centra evaluation device can evaluate the camera images of a plurality of cameras, wherein e.g. a plurality of vehicle seats can be classified simultaneously. For example, a central evaluation device contains model data of the plurality of vehicle seats.

A further development of the classification system further comprises a classification device, which is designed to output a classification result of the vehicle seat based on the camera image transferred by the evaluation device with the model data. The development of the classification system can allow higher automation in a classification of driving seat seats.

The disclosure also includes the further developments of the disclosed classification system having one or more features as already described in connection with the further developments of the disclosed method. For this reason, the corresponding Witer educations of the disclosed classification system are not described here again, but also apply to them as disclosed.

Exemplary embodiments are described below with reference to the accompanying figures tert erläu. Show it:

Fig. 1 is a schematic representation of a method for classifying a vehicle seat;

FIG. 2 shows a classification system with a standard weight arranged on a vehicle seat and an image evaluation device; FIG. and Fig. 3 is a further schematic representation of a method for classifying a vehicle seat, wherein the vehicle seat is associated with a quality level or a comfort class.

Various embodiments will now be described in more detail with reference to the accompanying drawings, in which some embodiments are shown. In the figures, the thickness dimensions of lines, layers and / or regions may be exaggerated for the sake of clarity.

1 illustrates a schematic representation of a method 10 for classifying a vehicle seat. The method comprises positioning at least one standard weight 11 on the vehicle seat, creating at least one camera image 12 of the vehicle seat with the standard weight, overlaying the camera image 13 with model data of the vehicle seat, the model data having classification ranges and classifying the vehicle seat 14 by means of evaluation of the model data overlaid camera image. The method makes it possible to classify vehicle seats quickly and inexpensively with camera images that are augmented by virtual elements.

2a shows an exemplary classification system 20 with a vehicle seat 21 arranged standard weight 22 and an image evaluation device 23. The Bildauswertungsvorrich device 23 may include a camera 24 and a computing unit 25, which may be an evaluation device and, for example, processor, memory, image analysis software and model data. The camera 24 is designed to len a camera holder 26 of the vehicle seat 21 erstel, wherein the camera holder 26 is a photo or an image recording. The camera mount 26 can also be a video. The vehicle seat 21 includes a seat surface 27, a seat back 28 and a headrest 29. The standard weight 22 of the classification system 20 may be positioned on the seat surface 27 of the vehicle seat 21.

In Fig. 2a, a cross-sectional view A of the seat surface 27 is shown. The cross-sectional view A shows side support pads 30, 30 'of the vehicle seat 21. In the cross-sectional view A, a virtual seat surface 31 and a real seat surface 32 are shown. The real seating surface 32 may be determined from the camera receptacle 26. The virtual seat surface 31 may be stored in CA (computer aided design) data of the classification system 20 superimposed with the camera shot. For example, the shape and position of the virtual seat surface 31 corresponds to a real seating surface of the vehicle seat 21 when no applied weight on the seat surface deforms it. The arithmetic unit 25 of the classification system 20 may be configured to superimpose the virtual seat surface 31 (the CA data) in the camera receptacle 26 with the real seating surface 32 in such a way that real and virtual seating surfaces may be overlaid. Surface when the seat surface is not weighted by a weight, or to overlap in undeformed places, such as the backrest when the seat cushion is deformed. Based on a comparison of the real seat surface 32, which may be deformed by the weight 22, for example, and the virtual seat surface 31, the arithmetic unit 25 can determine in example, with what strength or to what degree the seat cushion is deformed by the weight of Normge. Based on the determination of the deformation and the mass of the standard weight, the arithmetic unit can classify, for example, a strength of the seat cushion of the seat surface 27.

2b shows an enlargement 35 of the cross-section A of the vehicle seat 21. In Fig. 2b, the virtual seat surface 31 and deformed by the standard weight 22 real seat surface 32 are abgebil det. The standard weight 22 has a first measuring point 36, a second measuring point 37 and a third measuring point 38. The measuring points are measuring lines that can indicate classification areas. For example, a position of a measuring point on the camera image 26 in the United ratio to the virtual seat surface 31 for a classification of the strength of the seat cushion ver used.

In a first example, the seat cushion may be soft, ie have a low strength, so that it is greatly deformed by the weight 22. In this case, in the camera image 26, the first measurement point 36 may lie on a contour line of the virtual seat surface 31, and the vehicle seat 21 may be classified as a vehicle seat having a soft seat cushion. Thus, it can be recognized whether a strength of a seat cushion is too low compared to a standard, the standard being e.g. can be displayed by the second measuring point 27.

In a second example, the seat cushion of the seat surface 27 is hard so that the weight 22 less deforms the seat cushion. For example, a distance of the real seat surface 32 to the virtual seat surface 31 is smaller than in the first example. The measuring point 38 may be a measuring line which lies in the camera image 26 on the virtual seat surface 31. The measuring point 38 may be associated with a hard seat cushion or high strength of the seat cushion, so that when the measurement point 38 is on the virtual seat surface 31, the vehicle seat 21 or the seat cushion of the seat surface 27 of the vehicle seat 21 is considered hard, e.g. Harder than a standard, can be classified.

FIG. 3 shows a further schematic representation of a method 40 for classifying a vehicle seat, wherein a quality level and / or a comfort class can be assigned to the vehicle seat. The method 40 includes importing augmented reality for classifying vehicle seats, for example by overlaying CA data 41 of the vehicle seat in a co-location with the real seat. The real seat may be the vehicle seat received by the camera. The CA data may be a virtual design, such as a digital design model of the vehicle seat. Overlaying may include adapting the CA data to the captured image of the vehicle seat depending on the camera's perspective, for example, in terms of size, angle of rotation or translation angle, or adjusted for optical distortion of the camera. The method 40 includes, for example, an optical evaluation 42, which may be a qualitative assessment of the seat adjustment. For example, it can be checked whether the vehicle seat is set in the co-location.

The method 40 includes, for example, positioning the reference weight 43. It may further include overlaying a boundary layer 44 of the CA model with the captured image of the vehicle seat. For example, a limit position of the CA model describing a surface of a seat cushion of the vehicle seat is superimposed on the surface of the seat cushion of the vehicle seat on an image taken by the camera. By way of example, this makes it possible to check the position of the reference weight 45. In the CA data, an intended target position of the reference weight can be stored and an image recognition device can be designed to determine a position of the reference weight on the camera image. An optical adjustment 46 of the CA boundary layers with a deformation, for example a deformation of the seat cushion of the vehicle seat, by the reference weight can be performed. Finally, it is possible to classify in a comfort class 47. For example, the classification into the comfort class depends on the deformation of the seat cushion.

According to the method 40, it is also possible to superimpose a CA-KO layer with the real seat, for example the vehicle seat on a camera image. It can be done a opti shear boundary adjustment 49, in which based on the CA data, a boundary position adjustment (for example, tolerance levels) of about Kedem, filing or stitching of the vehicle seat Runaway leads can be. For example, it can be determined based on the camera image of the vehicle seat without a Refe rence weight by matching with the CA data and pattern recognition of the recorded Bil of whether boundary layers, such as seams of the vehicle seat, are attached to a designated position. The intended position can be stored in the CA data. As a result, a classification into quality level 50 is possible, for example if an actual limit position of the vehicle seat on the camera image deviates slightly, for example less than 5 mm, less than 3 mm or less than 1 mm, from the limit position of the superimposed CA data, a quality level can be high, eg the vehicle seat is assigned a highest value of a quality scale.

It is possible according to method 40 to record the evaluation basis, for example the classification into quality level 50 or the classification into comfort classes 47 by means of documentation 51. Beispielswei- The camera image can be stored with the superimposed CA data or an evaluation of the camera image can be documented, for example by what degree a seat cushion of the vehicle seat has been deformed by arranging the reference weight on the vehicle seat.

In summary, it is proposed in the present application, by means of a Kamerys system (eg in a tablet / Hololense / smartphone) and a software with deposited seat boundary pattern data and a hardware reference object (weight) perform a comparison of the deformations of a vehicle seat virtually and with hardware. By overlaying virtual CA data with a real seat, optical discrepancies are detected. About the reference weight on the seat and its generated deformation in the seat foam (sinking) is objectified and predicted over the optical from the same seating comfort. Furthermore, compliance with the comfort specification can be monitored in series operation / production without costly subject studies. The proposed method may allow for time saving and objectification in the classification of vehicle seats, as well as providing the possibility of rapid series monitoring in the manufacture of vehicles regarding the quality of the vehicle seats.

LIST OF REFERENCE NUMBERS

10 V learn to classify

11 Positioning of a standard weight

12 Creating a Camera Image

13 Overlay the camera image with model data

14 classifying the vehicle seat

20 classification system

21 vehicle seat

22 standard weight

23 image evaluation device

24 camera

25 arithmetic unit

26 picture recording

27 seat area

28 backrest

29 Headrest

30 side cushion

31 virtual seat surface

32 real seat surface

35 magnification

36 first measuring point

37 second measuring point

38 third measuring point

40 Methods for Classifying

41 Overlay CA data

42 Optical rating

43 Positioning the Reference Weight

44 Overlay Borderline

45 Checking the position of the reference weight

46 Optical adjustment

47 Classification in comfort class

48 Overlay CA-KO position

49 Optical boundary adjustment

50 classification in quality level

51 documentation

Claims

claims

A method (10) for classifying a vehicle seat (21) of a motor vehicle, the method (10) comprising:

Positioning at least one standard weight (11) on the vehicle seat (21);

Creating at least one camera image (12) of the vehicle seat (21) with the standard weight;

Overlaying the camera image with model data (13) of the vehicle seat (21), the model data having classification ranges; and

Classifying the vehicle seat (14) by means of evaluation of the superimposed with model data Kame rabildes.

2. The method (10) of claim 1, wherein the standard weight is positioned on a seat cushion of the vehicle seat (21).

3. The method (10) according to claim 1 or 2, wherein for superimposing the camera image the vehicle seat (21) associated with model data are used and the model data at least in formations to the unloaded vehicle seat.

4. The method (10) according to any one of the preceding claims, wherein the Klassifizierungsberei surface are assigned to a respective standard weight and a degree of plastic deformation of a surface (32) of the vehicle seat (21) detect by the respective standard weight.

5. The method (10) according to any one of the preceding claims, wherein at least three Klassifi be used zierungsbereiche, wherein compared to a standard strength of a vehicle seat surface a first classification range, a strength of the vehicle seat (21) is assigned higher than the standard strength, a second classification range one of Standardfestig strength is assigned corresponding strength and a Fes activity is assigned less than the standard strength of a third classification range.

A method (10) according to any one of the preceding claims, further comprising the method

Determining a measurement temperature and a measurement humidity when the camera image is created and normalizing a classification of the vehicle seat (14) to a standard temperature and a standard humidity using a material table comprising at least one material of the vehicle seat (21).

7. The method (10) according to any one of the preceding claims, wherein the standard weight has at least one measurement mark, which is designed for comparison with the model data and assigning a classification range.

8. The method (10) according to any one of the preceding claims, wherein the camera image is created after a period of time longer than 60 s after positioning the standard weight (11) on the driving convincing seat.

9. The method (10) according to any one of the preceding claims, wherein the standard weight has a mass of at least 3 kg and at most 20 kg.

10. The method (10) according to any one of the preceding claims, wherein the standard weight on a headrest, a backrest and / or a side support cushion of the vehicle seat (21) is posi tioned.

11. The method of claim 1, wherein two standard weights for classifying a vehicle seat are positioned at two different locations of the vehicle seat.

12. The method according to claim 1, wherein the classifying comprises checking a functionality of a movement device of the vehicle seat, wherein a movement of the standard weight caused by the activated movement device is determined by evaluating at least two camera images, based on the movement of the standard weight the functioning of the movement device is classified.

A classification system (20) for classifying a vehicle seat (21), comprising: at least one standard weight (22); and

An image evaluation device (23) comprising at least one camera (24) and an evaluation device (25), wherein the evaluation device (25) has model data for at least one vehicle seat (21) and is adapted to a camera image (26) of the camera (24). to overlap with the model data.

14. The classification system (20) of claim 13, further comprising

A classification device configured to output a classification result of the vehicle seat (21) based on the camera image (26) superimposed on the model data by the evaluation device (23).