• 专利附录
  • 专利说明
  • 权利要求
  • 类似技术
  • 同族专利
  • 引用文献
  • 法律状态
  • 优先权
    • 专利摘要:
    The invention relates to a method for measuring the position of a mobile structure in a repository, the structure comprising a plurality of recurring elements intrinsic to the structure, the method comprising the following successive steps: a) acquisition (20), by processing means, images of the structure taken simultaneously by a plurality of optical devices, each recurring element of the structure being in the field of view of at least three distinct optical devices; b) extraction (21a, 21b) on each of the images, by the processing means, of the recurring elements and determination (21c), by the processing means, of the position of said recurring elements in the image; c) calculating (22a) by, the processing means, at least one indicator for each of the recurrent elements detected on each image; d) identifying (22b), by the processing means, each of said recurring elements by associating a unique identifier with each of said elements, the identifier being linked to the position of each of the optical devices in the reference frame and to at least one indicator associated with each of said elements on each image; and e) determining the position (23), by the processing means implementing a photogrammetry algorithm, of each recurrent element in the repository.
    公开号:FR3036473A1
    申请号:FR1554526
    申请日:2015-05-20
    公开日:2016-11-25
    发明作者:Jean-Christophe Bry；Claude Leonetti；De Langautier Gregoire Pebernad
    申请人:Airbus Operations SAS；
    IPC主号:
    专利说明:

    [0001] The present invention relates to a method for measuring the position of a mobile structure, in particular a structure of an aircraft fuselage. An aircraft fuselage is, in known manner, constituted by the assembly of a plurality of structures, such as the nose, the rear portion or different cylindrical sections.
    [0002] The assembling step of two structures consists in positioning the structures on supports and in bringing together a structure mounted on a mobile support of the other structure mounted on a fixed support, until the two structures are joined. The structures are then permanently fixed to each other, for example, by riveting. During its displacement, the position of the mobile structure is constantly measured in order to ensure the smooth running of the assembly step. A photogrammetry method is used to perform these measurements. Such a method makes it possible to construct a three-dimensional point cloud of the mobile structure by acquiring, via several cameras, images at different angles of a plurality of targets distributed over the entire structure and using the parallax obtained between the different images acquired.
    [0003] The scatter plot gives the respective position of the structure in a given frame. Each target, as described in particular in the document US 20070153297, comprises an optical code specific to the target. The target is reflective when illuminated by a laser projector so as to be detectable by the cameras. The use of such targets for the measurement of the position of a structure slows the production rates because operators must install the plurality of targets (for example of the order of 200 for a forward tip) only for the measurements then the uninstall once done. There is therefore a need for a method of measuring a position of a moving structure that is faster while being accurate. The invention relates to a method for measuring the position of a mobile structure in a repository, the structure comprising a plurality of recurring elements intrinsic to the structure, the method comprising the following successive steps: a) acquisition, by means of processing, images of the structure taken simultaneously by a plurality of optical devices, each recurring element of the structure being in the field of view of at least three separate optical devices; b) extracting on each of the images, by the processing means, recurring elements and determining, by the processing means, the position of said recurring elements in the image; c) calculating, by the processing means, at least one indicator for each of the recurrent elements detected on each image; D) identification, by the processing means, of each of said recurring elements by associating a unique identifier with each of said elements, the identifier being linked to the position of each of the optical devices in the repository and to at least one associated indicator each of said elements on each image; and 5 e) determining the positio, by the processing means implementing a photogrammetry algorithm, of each recurrent element in the frame of reference. One of the advantages of the present invention is the use, as targets, of intrinsic recursive elements in a structure that makes it possible to determine the position of said structure in a given referential without requiring the temporary setting of targets. A photogrammetry measurement method can therefore be implemented without having to carry out specific prior preparation of the structure. The characteristics of the invention mentioned below, as well as others, will become more clearly apparent on reading the following description of exemplary embodiments, said description being given in relation to the accompanying drawings, among which: FIG. 1 is a view of the device for determining the position of a mobile structure installed in an assembly hall according to one embodiment of the invention, wherein the structure comprises a skin held by rivets; FIG. 1A is an enlarged view of the skin of the structure of FIG. 1 illustrating the configuration of the rivets on a portion of said skin; FIG. 2 is a diagram illustrating the four-step method implemented by the device of FIG. 1 for determining the position of the structure according to one embodiment of the invention; FIG. 3 is a view similar to FIG. 1 in which the device for determining the position of a structure is arranged to control the assembling operations of the mobile structure to a fixed structure. With reference to FIGS. 1 and 1A, a device 100 for measuring the position of a structure 30 is arranged so as to measure the position of a first structure 1 placed on a support 2. The first structure 1 is movable according to a horizontal translation axis X. For this purpose, the first structure is fixed to the support via slides 2a allowing the displacement in translation of the first structure 1 relative to the support 2. The first structure 1 is for example a fuselage section and is formed by structural elements of the frame and smooth type (not shown in the figures) forming a skeleton of the structure and on which are fixed panels 5, for example metal or composites 3036473 3, forming the skin 6 of the structure. The panels 5 are fixed to the skeleton by rivets 10. The rivets 10 are in known manner arranged in the form of vertical lines 12 (along the Z axis), horizontal 13 (along the X axis), or diagonal 15, crossing and where the gap between the rivets can be variable. In known manner, the rivet heads 11 are visible and can be distinguished from the skin 5. According to the invention, and as will be described later, the rivets 10, as recurring and intrinsic elements of the skin 6 of the first structure 1, are used as targets by the measuring device 100 to measure, by photogrammetry, the position of the first structure 1 with respect to the second structure 3. By intrinsic elements is meant indissociable elements / constituents of the first structure 1. The measuring device 100 comprises a plurality of optical devices 101, of the camera type, and processing means 102, of the central processing unit type, for the processing of the electrical signals coming from the optical devices 101 to which the processing means are connected.
    [0004] The number of optical devices 101 is adapted to take images of the entire skin 6 of the first structure 1 such that each rivet 10 of said structure is in the field of view of at least three optical devices 101. In order to measure the position of the first structure 1 by photogrammetry along its axis of travel X, the optical devices 101 are further arranged to cover a predetermined path of the first movable structure 1. For this purpose, the optical devices 101 are positioned at a predetermined distance, for example of the order of 1 to 2 meters, from the skin 6 of the first structure 1. The optical devices 101, for example of the CMOS camera type (semiconductor complementary metal-oxide conductors) or CCD (charge-coupled device), have a definition of at least 25 megapixels per camera. The spatial position of the cameras 101 in a fixed measurement frame O, X, Y, Z called common reference is recorded in a memory of the processing means 102. In the example shown in FIG. are, for example, fixed to the ceiling (not shown) of an assembly hall 110. The common reference system O, X, Y, Z is in this case related to the assembly hall 110. According to the invention, and with reference to FIG. 2, the processing means 102 are configured to implement an image acquisition step taken simultaneously by each of the optical devices 101 and to implement different image processing algorithms for: in a recognition step 21, extract the rivets 10 from the other elements of the skin 6 of the first structure 1 and determine the position of these rivets on each image, - in an identification step 22, identify each rivet 10 detected on different images, and - in a recognition step 23 of the position of each rivet, determine the respective position of each rivet identified in the common reference system via a photogrammetry process. The purpose of the image processing implemented by the processing means 106 is the construction of a cloud of points, the points being the centers 11 of the detected rivets, in three dimensions of the first structure 1 in the common reference frame O, X, Y, Z to measure the position of the latter. In the step of recognizing the rivets 21, the processing means 102 implement a recognition algorithm applied to each image, worked in gray levels. This algorithm consists of an extraction operation 21a, 21b of the rivets of the image and a determination operation 21c of the coordinates of the centers of the rivets extracted. The extraction operation 21a, 21b of the rivets makes it possible to separate the rivets 10 from other elements present on the shot, such as the skin 6 or surface asperities of the latter. For this purpose, this operation comprises the application of two successive filters which are applied to each of the images: an adaptive thresholding filter 21a so as to detect a characteristic shape of the rivets 10, such as, for example, the rivet heads 11. In the image displayed in gray level, the head of a rivet being white (respectively black), all detected points of gray level too low (respectively too high) are removed; and a contour detection filter 21b of the heads 11 applied to the previously filtered image to enhance the detection of the contours of the heads so as to accurately locate their extent on the image. For this purpose, a morphological gradient contouring filtering technique is employed, which technique includes diluting the filtered image and eroding the filtered image, and subtracting the eroded image from the image. dilated image.
    [0005] Once these two filters are applied, the image comprises only surfaces closed by contours that may have disparate shapes so that the precise coordinates of the rivets on the image, and in particular their centers, can not be determined. . These disparate forms are notably due to poor illumination conditions of the heads, reflections on the heads, or heads deformed by the curvature of the skin.
    [0006] The determination operation 21c of the coordinates of the extracted rivets 10 is carried out in order to accurately detect the coordinates of the rivets extracted on the image. This operation 3036473 5 consists in the search, for each contour, of the smallest ellipse containing the contour. At the end of this operation, the coordinates of the center of the ellipses, which is similar to the center of the rivets, and the coordinates of the ellipses are determined on each image.
    [0007] In a second step, the processing means implement the identification step 22. This step comprises the implementation of an identification algorithm which consists, in a first step, in a calculation operation 22a. at least one indicator for each ellipse of each image. The indicators are related to the configuration (vertical lines, horizontal lines, crossing ...) of the rivets 10, identified as ellipses in the image, on the skin of the first structure. The indicators are, for example, taken from among the following indicators: the number of ellipses in the vicinity of the studied ellipse; if the ellipse is included on a horizontal line of ellipse; if the ellipse is included on a vertical line of ellipse; if the ellipse is included on a diagonal line of ellipse; 15 - if the ellipse is included is at the intersection of two lines; - if the ellipse is at the center of a cross of vertical and horizontal ellipse lines; - if the ellipse is at the end of the ellipse line; the determination of the angle of incidence with the centers of the nearest ellipses; - the position of the ellipse in the image.
    [0008] The identification algorithm then consists, in a second step, in an operation of identification of each of the ellipses by an identifier of its own. For this purpose, the processing means 102 implement a learning algorithm of the class of the wide margin separators (in English: Support Vector Machine) taking as input data the position of each of the optical devices 101 in the common reference frame O, X, Y, Z and the indicator or indicators associated with each ellipse on the images. This algorithm, aimed at image registration, consists of pairing for a pair of images taken by two different optical devices 101, the ellipses which have the same indicators on the first and on the second image. At the end of the identification step 22, each ellipse, which is found on at least three images taken by three different cameras, is associated with a specific identifier (label). The coordinates of each center of an identified ellipse are thus determined on at least three different images. Once the ellipses have been identified with a unique identifier, the processing means 35 implement the step of determining the position of each rivet 23. This step comprises the implementation of a photogrammetry algorithm 23 for determining the coordinates of each center of ellipse identified in the common reference frame O, X, Y, Z. Such an algorithm, based on the principle of triangulation, is known to those skilled in the art and therefore will not be detailed further. The purpose of step 23 is to construct a three-dimensional point cloud of the first structure 1, the points being the centers of the ellipses (substantially of the same coordinates as the centers of the rivets). This point cloud makes it possible to give the respective position of each rivet 10 in the common reference frame, and thus to determine the position of the first structure 1 in the common reference frame O, X, Y, Z Advantageously, the measuring device 100 according to the The invention may be connected to a display device comprising a screen for displaying on the latter the position of the points in the common reference frame O, X, Y, Z. Thus, operators can directly monitor and control the movement of the first structure. The invention is advantageous in that the point cloud is constructed without the use of external targets. Indeed, according to the invention, it is the rivets 10 already present in the structure which are used as targets for photogrammetry. Thus, the invention eliminates the steps of installation and removal of targets, and thus meets the aforementioned need for a position measuring means of a structure that is quick to implement 20 In addition to this advantage , the measuring device 100 according to the invention can be installed permanently in an assembly hall 110 unlike the laser projector used in measurement techniques according to the prior art. In addition, the use of a laser requires special precautions for use and therefore, the device according to the invention which uses only camera-type optical devices is therefore simpler to implement.
    [0009] Without departing from the scope of the present invention, the measuring device 100 as described can be used to quickly and simply produce three-dimensional digital models of a structure comprising rivets. For this purpose, the positions of the rivets are recorded in a database written in a memory of the measuring device 100.
    [0010] With reference to FIG. 3, the measuring device 100 according to the invention is used for the automation of the assembly of the first structure 1 to a second structure 3. The second structure 3 is mounted on a fixed support 30 and is, for example, another fuselage section formed identically to the first structure.
    [0011] The step of assembling the two structures 1, 3 is to bring, by translation, along the horizontal axis X, the first structure 1 mounted on the mobile support 2 of the second structure 3 3036473 7 until the two structures are joined, they are then permanently fixed to one another for example, by riveting. The mobile support 2 comprises means for translating the first structure 1, for example motorized slides 2a whose movement is controlled by a motor 5 driven by the processing means 102. In addition, the first structure is fixed to the movable support via actuators 2b, of the jack type, controlled by the processing means 102. The processing means 102 are provided for comparing, at regular intervals, the successive positions along the horizontal axis X of the first structure 1 measured by the means of processes 102 according to the method described above and predefined positions stored in a memory of the processing means. If there is a difference between the two compared positions, the processing means 102 control the actuators 2b and / or the motor 2a of the slides 2a to correct the position of the first structure 1. This embodiment makes it possible to make automatic corrections. the alignment of the two structures 1.3 so as to respect the assembly and assembly tolerances of the first structure 1 to the second structure 3. This embodiment is advantageous in that the measurement of the position of the first structure does not require the setting of targets on the latter. Thus, the implementation of this embodiment is simple and fast. The invention has been described using a rivet 10 as a recurring and intrinsic element of a structure as a target for employing a photogrammetry method. Without departing from the scope of the present invention, other recurring and intrinsic elements of a structure, such as weld points (for example automatic industrial MIG welding points), or bolts or screws could be used in place and place of rivets.
    权利要求:
    Claims (7)
    [0001]
    CLAIMS1) A method for measuring the position of a structure (1) in a mobile reference (O, X, Y, Z), the structure comprising a plurality of recurring elements (10) intrinsic to the structure, the method comprising the successive steps: a) acquisition (20), by processing means (102), of images of the structure (1) taken simultaneously by a plurality of optical devices (101), each recurrent element (10) of the structure being in the field of view of at least three separate optical devices; b) extraction (21a, 21b) on each of the images, by the processing means, of the recurring elements and determination (21c), by the processing means, of the position of said recurring elements in the image; c) calculating (22a) by, the processing means, at least one indicator for each of the recurrent elements detected on each image; d) identifying (22b), by the processing means, each of said recurring elements by associating a unique identifier with each of said elements, the identifier being linked to the position of each of the optical devices in the reference frame and to at least one indicator associated with each of said elements on each image; and e) determining the position (23), by the processing means implementing a photogrammetry algorithm, of each recurrent element in the reference frame (O, X, Y, Z).
    [0002]
    2) Method according to claim 1, characterized in that step b) comprises: - an adaptive thresholding (20a) of each of the images so as to detect a characteristic shape (11) of the recurring elements; and - a detection operation (20b) of a contour for each detected characteristic shape.
    [0003]
    3) Method according to claim 2, characterized in that step b) comprises the search (21c), for each contour, the smallest ellipse containing the contour.
    [0004]
    4) Method according to any one of claims 1 to 3, characterized in that each indicator is calculated (22a) according to the configuration of the recurring elements (10) on the structure (1), and is taken among the indicators next: - the number of ellipses in the vicinity of the studied ellipse; if the ellipse is included on a horizontal line of ellipse; if the ellipse is included on a vertical line of ellipse; 3036473 9 - if the ellipse is included on a diagonal line of ellipse; if the ellipse is understood is at the intersection of two lines of ellipse; - if the ellipse is at the center of a cross of vertical and horizontal ellipse lines; - if the ellipse is at the end of the ellipse line;
    [0005]
    The determination of the angle of incidence with the centers of the nearest ellipses; - the position of the ellipse in the image. 5) The method according to claim 3, characterized in that step d) comprises implementing a learning algorithm of the class of wide margin separators so as to pair (22b), for a first and a second time. second images taken by two separate optical devices (10 there, the ellipses which have the same indicators on the first and on the second image.
    [0006]
    6) Method according to any one of the preceding claims, comprising an additional step of displaying, on a screen, the position of each recurring element (10) of the structure in the repository (O, X, Y, Z).
    [0007]
    7) Process according to any one of the preceding claims, characterized in that the intrinsic recurring elements (10) to the structure are taken from rivets, bolts, weld points.
    类似技术:
    公开号 | 公开日 | 专利标题
    FR3036473A1|2016-11-25|METHOD FOR MEASURING THE POSITION OF A MOBILE STRUCTURE
    FR3037429A1|2016-12-16|SYSTEM AND METHOD FOR AUTOMATIC SURFACE INSPECTION
    US10302424B2|2019-05-28|Motion contrast depth scanning
    EP3782121A1|2021-02-24|Multistage camera calibration
    CA2891149A1|2015-11-27|Method for projecting virtual data and device allowing this projection
    CN103604808A|2014-02-26|Visual detection method for bottle cap defects
    FR3025918A1|2016-03-18|METHOD AND SYSTEM FOR AUTOMATED MODELING OF A PART
    EP3072109A1|2016-09-28|Method of estimating the speed of displacement of a camera
    WO2015071457A1|2015-05-21|Method for estimating the speed of movement of a camera
    EP3271869B1|2019-10-16|Method for processing an asynchronous signal
    CA2880145C|2020-07-28|Method for the non-destructive testing of a blade preform
    EP3272119A1|2018-01-24|Method for the 3d reconstruction of a scene
    EP3070643B1|2021-09-29|Method and device for object recognition by analysis of digital image signals representative of a scene
    KR101238748B1|2013-03-04|System for measuring distance of target using step-staring infrared sensor unit
    CN108550160B|2020-04-07|Non-uniform light bar characteristic region extraction method based on light intensity template
    EP0329224B1|1992-01-15|Process for measuring lengths using a photosensitive array camera
    EP2862154B1|2020-11-18|Grouping of data attached to images
    EP0604252B1|1997-05-14|Method for aiding piloting of a low flying aircraft
    FR3025025A1|2016-02-26|METHOD FOR DETERMINING THE RELATIVE POSTURE OF A MOBILE MACHINE WITH RESPECT TO AN ART WORK AND ADAPTIVE MOBILE MACHINE
    EP3301644B1|2020-10-21|Method for building a depth map of a scene and/or an all-in-focus image
    FR3096816A1|2020-12-04|Method, system and apparatus for the detection of product facades
    FR3103301A1|2021-05-21|Method for detecting specularly reflected light beam intensity peaks
    EP3182374A1|2017-06-21|Automated method for three-dimensional optical measurement
    FR3015033A1|2015-06-19|METHOD AND DEVICE FOR ANALYZING THE SURFACE OF A SUBSTRATE
    WO2018141717A1|2018-08-09|Method for identifying and monitoring areas of interest in images successively acquired by an optical sensor of a motor vehicle
    同族专利:
    公开号 | 公开日
    FR3036473B1|2018-11-16|
    US10124877B2|2018-11-13|
    US20160340016A1|2016-11-24|
    引用文献:
    公开号 | 申请日 | 公开日 | 申请人 | 专利标题
    US20120257017A1|2007-08-10|2012-10-11|Leiga Geosystems Ag|Method and surveying system for noncontact coordinate measurement on an object surface|
    US20120133739A1|2010-11-30|2012-05-31|Fuji Jukogyo Kabushiki Kaisha|Image processing apparatus|
    US20140105457A1|2011-07-05|2014-04-17|Hexagon Technology Center Gmbh|Method for providing target point candidates for selecting a target point|
    US20070153297A1|2006-01-04|2007-07-05|Lau Kam C|Photogrammetric Targets|
    US7783376B2|2006-05-10|2010-08-24|The Boeing Company|Photogrammetric contrasting light for hole recognition|
    US7454265B2|2006-05-10|2008-11-18|The Boeing Company|Laser and Photogrammetry merged process|
    DE102010010685A1|2009-03-19|2011-02-03|Airbus Operations Gmbh|Method for tolerance-adapted adhesive application in vehicle construction|
    US9377778B2|2010-02-17|2016-06-28|The Boeing Company|Integration of manufacturing control functions using a multi-functional vision system|
    CA3025538A1|2012-03-02|2013-09-06|Sight Machine, Inc.|Machine-vision system and method for remote quality inspection of a product|
    FR3052577A1|2016-06-13|2017-12-15|Airbus Operations Sas|METHOD FOR ADJUSTING THE POSITIONING OF ELEMENTS OF A STRUCTURE OF AN AIRCRAFT AND DEVICE FOR IMPLEMENTING IT|CN108372935B|2016-12-21|2021-09-10|中国航空制造技术研究院|Posture adjusting and docking system and method for large-range movement and automatic navigation|
    CN108557111B|2018-04-28|2020-07-28|成都致冠科技有限公司|Coordination assembly method for unmanned fixed-wing aircraft body|
    US20200017238A1|2018-07-12|2020-01-16|The Boeing Company|Wing panel assembly system and method|
    WO2021099648A1|2019-11-18|2021-05-27|Asociacion Centro Tecnologico Ceit|Dimensional control system based on photogrammetric vision|
    法律状态:
    2016-05-20| PLFP| Fee payment|Year of fee payment: 2 |
    2016-11-25| PLSC| Publication of the preliminary search report|Effective date: 20161125 |
    2017-05-23| PLFP| Fee payment|Year of fee payment: 3 |
    2018-05-22| PLFP| Fee payment|Year of fee payment: 4 |
    2019-05-23| PLFP| Fee payment|Year of fee payment: 5 |
    2020-05-22| PLFP| Fee payment|Year of fee payment: 6 |
    2021-05-20| PLFP| Fee payment|Year of fee payment: 7 |
    优先权:
    申请号 | 申请日 | 专利标题
    FR1554526A|FR3036473B1|2015-05-20|2015-05-20|METHOD FOR MEASURING THE POSITION OF A MOBILE STRUCTURE|
    FR1554526|2015-05-20|FR1554526A| FR3036473B1|2015-05-20|2015-05-20|METHOD FOR MEASURING THE POSITION OF A MOBILE STRUCTURE|
    US15/159,240| US10124877B2|2015-05-20|2016-05-19|Method for measuring the position of a mobile structure|
    [返回顶部]