Equipment for defining motion sequences on generic models
By adjusting the motion sequence on a general model, the problem of complex motion sequence definition in robot operations on unknown surfaces is solved, and accurate and efficient motion sequence definition is achieved, which is suitable for a variety of application scenarios.
Patent Information
- Application Number
- CN202080074843.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Priority Date
- 2019-12-10
- Filing Date
- 2020-11-30
- Publication Date
- 2024-11-08
- Estimated Expiration
- 2040-11-30
AI Technical Summary
Prior art When defining and adjusting motion sequences, especially in robots operating on unknown surfaces, requires a lot of computing power and precise motion management, resulting in complex and time-consuming definition processes, especially for massage robots and handmade porcelain painting robots. In other words, the movement must be very precise.
By obtaining the motion of reference elements that evolved on the surface of variable geometric shapes and adjusting the sequence of motion on the general model, two methods are adopted: one is to transform the general model to fit the object and then restore the initial parameters; the other is to stand alone The motion sequence is obtained from the general model and the difference between the object and the general model is calculated to transform the actual motion sequence into a general motion sequence.
It realizes the practical definition of motion sequences without the need for a three-dimensional representation or general model of visual surfaces, simplifies the definition process, improves accuracy and efficiency, can capture multiple motion sequences and adapt to different surfaces, suitable for a variety of applications Such as massage and handmade porcelain painting.
Smart Images

Figure CN114599487B_ABST
Abstract
Description
Technical Field
[0001] The invention relates to the field of defining at least one motion sequence on a generic model, ie on a digitized model with preset parameters to represent a set of objects.
[0002] The invention can be applied in many technical fields in which the working surface is variable between a plurality of individuals within a group of objects. Typically, the invention finds particularly advantageous application for defining one or more massage trajectories on a general human body. Alternatively, the individuals in the group of objects may correspond to physical objects. For example, the group of objects may correspond to porcelain plates, and the invention may be applied to define painting trajectories on these plates.
[0003] This motion sequence using reference elements on the generic model is also variable. In fact, by recording several motion sequences related to different objects and defining them on the generic model, it is possible to compare two actions performed on different objects. Furthermore, this motion sequence defined using the generic model can also be used to control a robot comprising means for adapting the generic motion sequence to a specific object. Background Art
[0004] Unlike industrial robots that are programmed to follow a preset trajectory, a robot that operates on an unknown surface must include a device to manage movement that can analyze the surface to determine a trajectory.
[0005] For example, a reconnaissance robot usually integrates at least one camera and image processing means which analyze the detected surface over time and determine the trajectory to be followed by the robot.
[0006] This method of analyzing unknown surfaces requires a lot of computing power to accurately guide the robot's movement over time. Therefore, the scout robot should move slowly to allow the motion management device to optimize the robot's movement based on the information acquired by the camera and processed by the image processing device.
[0007] Furthermore, for massage robots or hand-painted porcelain robots, the robot’s movements must be extremely precise in order to massage the desired areas of an individual’s body or apply layers of paint to the desired locations.
[0008] For this purpose, known processes for scanning three-dimensional surfaces enable the operator to program the movements of the robot by using a three-dimensional digital modeling of the surface to be treated by the robot. For example, document WO 2015 / 187092 describes a massage robot comprising a three-dimensional scanner for scanning a patient's body and allowing the practitioner to determine the massage trajectory of the robot using a projection of a three-dimensional model of the patient's body on a touch pad.
[0009] It is also possible to use a robot with at least one generic model for which the sequence of movements is known. The robot can then adapt the generic model to the object and simultaneously adapt the sequence of movements in order to perform treatment for the specific object.
[0010] For this type of robot, it is also necessary to define at least one motion sequence on the generic model. For the definition given in document WO 2015 / 187092, the definition of at least one motion sequence on the generic model is usually performed manually by an operator tracing a drawing on the generic model.
[0011] Therefore, due to the accuracy sought, the definition of this movement sequence may be particularly long.
[0012] The technical problem of the present invention consists in facilitating the process of defining at least one movement sequence on a generic model. Summary of the invention
[0013] The present invention aims to address this technical problem by acquiring the motion of a reference element evolving on a surface of variable geometry and by adapting the motion captured on a generic model.
[0014] Within the meaning of the present invention, a "reference element" may correspond to a manipulator or to a point or a set of reference points of a physical element. For example, a reference element may correspond to a glove or a set of points representing the practitioner's hand. Two methods may be used to adapt the actual motion to the generic model.
[0015] The first method consists in transforming the generic model to fit the object before acquiring the motion sequence. The acquisition of the motion sequence is then recorded directly on the transformed generic model. When the motion sequence is completed, the generic model is transformed again to retrieve these initial parameters, and then the motion sequence is transformed by retrieving the same transformation as applied to the generic model, thereby transforming the actual motion sequence into the generic motion sequence.
[0016] The second method is to obtain the motion sequence independently of the generic model and calculate the difference between the object and the generic model to apply the difference to the motion sequence to transform the actual motion sequence into the generic motion sequence. However, the second method requires the generic motion sequence to be correctly repositioned on the generic model.
[0017] The second method is applicable only to transformations of the translation, rotation and scaling type, while the first method can be applied to other types of transformations. In both cases, the invention relates to a device for defining a generic sequence of motions on a generic model, the device comprising:
[0018] - means for acquiring the position of a reference element moving on a surface; said reference element being configured to execute the actual movement sequence;
[0019] - means for recording said actual movement sequence;
[0020] - means for obtaining a three-dimensional representation of said surface;
[0021] - means for adapting said generic model to said three-dimensional representation of said surface; and
[0022] - means for defining a generic movement sequence on said generic model by applying said adjustment between said generic model and said three-dimensional representation of said surface to said actual movement sequence.
[0023] In the first case, the adjustment device is configured to adjust the generic model to adapt to the three-dimensional representation of the surface; the recording device is configured to record the actual motion sequence on the generic model while it is adapted to the dimensions of the three-dimensional representation of the surface; and the definition device is configured to transform the generic model using the recorded motion sequence so that the generic model restores these initial parameters.
[0024] In a second case, the adjustment means are configured to calculate the difference between the generic model and the three-dimensional representation of the surface; the recording means are configured to record the actual motion sequence independently of the generic model; the definition means are configured to transform the recorded motion sequence according to the difference calculated by the adjustment means, and the device includes means for positioning the generic motion sequence on the generic model.
[0025] Regardless of the adaptation method used, the present invention allows the motion sequence to be practically defined on the generic model without the need for a three-dimensional representation of the visual surface or the generic model.
[0026] Thus, multiple motion sequences can be captured and adapted to fit the same general model to compare multiple trajectories executed on the same surface or multiple different surfaces.
[0027] In the example of handmade porcelain painting, the present invention can therefore know the actions that the painter most often performs. In addition, by examining different renderings, it can be determined which actions are most effective.
[0028] Likewise, in the example of tracking of massage trajectories, it is possible to know the points most frequently requested by the practitioner, or to record one or more massage trajectories so that the robot can reproduce them. These massage trajectories can also be recorded on different individuals. In addition, multiple motion sequences can be digitized to perform several different types of massage. It is also possible to create several generic models to improve the adaptation of the generic model and the patient's body, for example, by using three generic models for each gender (large, small and medium body), for different age types (children, teenagers and adults) and / or for each recorded posture (sitting, standing and lying posture).
[0029] According to one embodiment, said means for acquiring the position of a reference element are configured to pick up the orientation of said reference element in order to transfer this orientation to different points of said generic motion sequence.
[0030] According to one embodiment, said means for acquiring the position of a reference element are configured to pick up actions performed or constraints suffered by said reference element in order to report these actions or these constraints at different points of said generic motion sequence.
[0031] This embodiment makes it possible, for example, to control the operation of the actuator during the movement of the robot. In the example of a handmade porcelain painting robot, the robot can perform specific surface treatments in certain places. In the example of a massage robot, certain positions of the robot can control a device that triggers vibrations to improve the comfort and / or effectiveness of the massage. In addition, a universal movement sequence can thus include several trajectories performed by a palpation-rolling movement, while other movements can be performed by another type of movement. Thus, the stresses experienced by the reference element can correspond to physical stresses (such as pressure or temperature) or external stresses (such as sounds emitted during a massage or the diffusion of essential oils).
[0032] According to one embodiment, the generic model and the three-dimensional representation of the surface are formatted into the shape of the point cloud, and the adjusting means comprises:
[0033] - means for calculating the direction of the normal to each point of said three-dimensional representation of said surface; and
[0034] - means for searching, for each point of the cloud of points of the three-dimensional representation, a point of a neighbouring generic model having a minimum difference between the direction of the normal of the point of the generic model and the direction of the normal of the point of interest;
[0035] - means for determining the distance between said detection point and said point of interest of said generic model; and
[0036] - means for searching for a global transformation of said generic model as a function of said distances determined for all point cloud points of said three-dimensional representation.
[0037] The normal direction may obtain information about the orientation of the surface of the generic model and the three-dimensional representation of the surface.Different from simple point-to-point coordinate comparison, surface comparison may provide more efficient recognition.
[0038] Furthermore, the adaptation of the generic model or the movement sequence is performed step by step by modifying the generic model or the movement sequence little by little, depending on the average value of the distances.
[0039] This embodiment thus makes it possible to efficiently adjust a generic model or a motion sequence by comparing the normal direction of each point of the generic model or motion sequence with the normal direction of the three-dimensional representation of the surface.
[0040] According to one embodiment, the searching means is configured to search for points of the generic model in a preset sphere surrounding the point of interest.
[0041] This embodiment aims to limit the search area of the general model points to limit the calculation time. In addition, the restriction of the search area also makes it possible to limit the modification range of the general model between two comparisons, thereby improving the modification accuracy of the general model.
[0042] According to one embodiment, the normal direction is determined by constructing the surface using the coordinates of the three or four points closest to the point of interest.
[0043] This embodiment allows efficient construction of surfaces and three-dimensional representations of surfaces of a generic model.
[0044] According to one embodiment, the adjustment device comprises:
[0045] - means for detecting feature points on said three-dimensional representation; and
[0046] - means for transforming the generic model in rotation and / or translation so that said positions of said characteristic points correspond to the positions of characteristic points of said generic model.
[0047] This embodiment allows to create a first rough adjustment of the generic model to increase the speed of the fine adjustment performed by the normals.In the example of a handmade porcelain painting robot, the feature points may correspond to the upper and lower ends of the porcelain.
[0048] In the example of the massage robot, the feature points may correspond to the upper end of the skull, the position of the armpits, and the position of the crotch.
[0049] According to one embodiment, the acquisition means comprises means for preprocessing the three-dimensional representation by capturing a plurality of three-dimensional representations and averaging the coordinates of points between the different three-dimensional representations. This embodiment makes it possible to improve the accuracy of the three-dimensional representation and therefore the fitness of the general model.
[0050] According to one embodiment, the preprocessing means filters the average value of the coordinates of the points between the different three-dimensional representations. This embodiment can also improve the accuracy of the three-dimensional representation and thus the fitness of the general model.
[0051] According to one embodiment, the reference element corresponds to a glove and the movement sequence corresponds to the movements made by the glove during a massage. BRIEF DESCRIPTION OF THE DRAWINGS
[0052] The embodiments of the present invention and the advantages thereof can be clearly understood from the following examples, which are provided for reference only and not for limitation. The examples are given by way of indication and not limitation. Figures 1 to 3 , which is composed of:
[0053] Figure 1 : A flowchart of the steps of determining the transformation of the general model according to one embodiment of the present invention;
[0054] Figure 2 : a flowchart of the operating steps of the apparatus for defining a generic motion sequence on a generic model according to the first embodiment; and
[0055] Figure 3 : Flowchart of the operating steps of the device for defining a general motion sequence on a general model according to the second embodiment.
[0056] In the following description, the invention is described with reference to the definition of a massage sequence. However, the invention is not limited to this specific application and it can be used for various motion sequences associated with surfaces whose geometry is not preset. DETAILED DESCRIPTION
[0057] like Figure 1 As shown, the surface analysis is performed by acquisition means 14 capable of providing a three-dimensional representation Re of the surface. The three-dimensional representation Re takes the form of a cloud of points, each point having three coordinates of an orthogonal system: x, y and z.
[0058] The acquisition means 14 may correspond to a set of photographic sensors, a set of infrared sensors, a tomographic sensor, a stereo sensor or any other known sensor capable of acquiring a three-dimensional representation of a surface. of The camera is used to obtain this three-dimensional representation Re.
[0059] To obtain this 3D representation Re without capturing the environment, a first point cloud corresponding to the environment and a second point cloud corresponding to the surface in its environment can be captured. Only the differences between the two point clouds are retained to extract the points corresponding to the surface from the environment. This approach can be abstracted from standardized environments for recording and adapted to any environment.
[0060] like Figure 1 As shown, these sensors 14 are usually implemented with pre-processing means 15 to provide a three-dimensional representation Re with improved quality or accuracy. For example, the pre-processing means 15 may correspond to an algorithm for equalizing histograms, filtering, averaging representations over a plurality of consecutive representations, etc.
[0061] For example, the method described in the scientific publication "KinectFusion: Real-time 3DReconstruction and Interaction Using a Moving Depth Camera*" published on October 16, 2011 at UIST'11 can be used to obtain a three-dimensional representation with better quality. The device then performs computer processing to adjust the general model m1, m2, m3 with the three-dimensional representation Re. The general model m1, m2, m3 is also formatted in the shape of a point cloud, where each point has three coordinates of an orthogonal system: x, y and z. Preferably, the general model includes an average model ModMoy of N vertices of three coordinates and a transformation matrix ModSigma of M morphological components consisting of 3N coordinates (i.e., three coordinates of N vertices). A large number of different people (e.g., one thousand people) are required to enrich each general model m1, m2, m3.
[0062] The data were subjected to dimensionality reduction using principal component analysis. By applying principal component analysis to these data, the variance in the data can be determined and the common variance can be associated with a component. Thus, instead of retaining one component for each person, each general model m1, m2, m3 stores about 20 components, explaining most of the variance for these thousand people. The method is described in more detail in the scientific publication "Building Statistical Shape Spaces for 3D Human Modeling, Pishchulin et al.", published on March 19, 2015 in the journal "Pattern Recognition 2017".
[0063] Preferably, the generic model m1 , m2, m3 is stored in a memory accessible by image processing means of the device capable of adapting the generic model m1 , m2, m3 and the three-dimensional representation Re.
[0064] To this end, when a three-dimensional representation Re is obtained, the device performs a detection of characteristic points Pref of this three-dimensional representation Re by means of digital processing means 16. Figure 1 In the example of , the feature points Pref correspond to the upper end of the skull, the position of the armpits and the position of the crotch. The digital processing device 16 can perform all known methods for detecting elements in an image (eg the Viola and Jones method).
[0065] Preferably, in order to detect feature points Pref, the point cloud is transformed into a depth image (i.e., a grayscale image), for example 12-bit encoding, so that a depth range of 0 to 4095 mm can be encoded. The depth image is then thresholded and binarized to highlight only pixels with a value of 1 corresponding to the object / body of interest and pixels with a value of 0 corresponding to the environment. Next, the binarized image is edge detected using a method disclosed, for example, in Suzuki, S. and Abe, K., "Topological Structural Analysis of Digitized Binary Images by Border Following", CVGIP 30 1, pp. 32-46 (1985). Finally, the convex points of the contour and their convexity defects (determined, for example, using the method of Sklansky, J., "Finding the Convex Hull of a Simple Polygon", RL1$number, pp. 79-83 (1982)) are used as feature points Pref.
[0066] The means 17 for selecting a generic model m1 , m2 , m3 are then implemented to select the generic model m1 , m2 , m3 that is closest to the three-dimensional representation Re.
[0067] For example, the height dimension of the three-dimensional representation Re can be roughly estimated by calculating the distance between the feature point Pref of the top of the skull and the feature point of the crotch and performing such selection by selecting the generic model m1, m2, m3 with the closest height dimension. Similarly, the selection of generic models m1, m2, m3 can be performed by using the width of the three-dimensional representation Re by calculating the distance between the armpit feature points Pref.
[0068] Furthermore, the generic models m1, m2, m3 can be articulated by virtual bones representing the most important bones in the human skeleton. For example, fifteen virtual bones can be modeled on the generic model m1, m2, m3 to define the position and shape of the spine, femur, tibia, ulna, humerus and skull. Furthermore, the orientation of these virtual bones makes it possible to define the pose of the generic model, i.e., if the generic model m1, m2, m3 has one arm in the air, legs spread apart...
[0069] The selection can also be performed by the device 17 to determine the pose of the generic model m1, m2, m3 by comparing the distance into the depth image contour of the object / body of interest with a database of depth image contours of generic models with thousands of poses (e.g., calculated using the Hu method, Visual Pattern Recognition by Moment Invariants, IRE Transactions on Information Theory, 8:2, pp. 179-187, (1962)). The depth image of the articulated generic model m1, m2, m3 that is closest to the depth image of the object / body of interest is selected, and the rotation value of the virtual skeleton is saved.
[0070] A first adjustment is then performed by the adjustment means 18 by transforming the selected generic model to approximate the three-dimensional representation Re. For example, the first adjustment may simply transform the width and height of the selected generic model so that the spacing of the feature points Pref of the selected generic model corresponds to the spacing of the feature points Pref of the three-dimensional representation Re. The first adjustment may also define the pose of the virtual skeleton of the generic models m1, m2, m3.
[0071] After this first rather rough adjustment, a more precise second adjustment can be used by using the normal directions formed by each surface defined between the points of the three-dimensional representation Re. To this end, the device comprises means 19 for calculating the normals of each surface of the three-dimensional representation Re and of the chosen generic model.
[0072] For example, the normal direction may be determined by constructing each surface of the three-dimensional representation Re using the coordinates of the three or four points closest to the point of interest. As a variant, the normal direction of the generic model may be calculated during the step of defining the generic model.
[0073] The device then uses a search device 20 which is capable of detecting, for each point of the point cloud of the three-dimensional representation Re, a point of a generic model selected in its vicinity, where the difference between the normal direction of the point of the generic model and the normal direction of the point of interest is minimal. When the virtual skeleton is a component of the selected generic model, the search device 20 adjusts the position and size of the virtual skeleton by changing the characteristics of each virtual skeleton to adjust the virtual skeleton to the position of the body element present in the three-dimensional representation Re.
[0074] For example, the search means 20 may be configured to search for points of the generic model in a preset sphere around the point of interest. Preferably, the radius of the sphere is determined according to the number of vertices of the generic model and the size of the object / body of interest, so that about 10 points are included in the sphere.
[0075] Using all these normal directions, the device can then calculate the difference between the selected generic model and the three-dimensional representation Re using determination means 21 capable of calculating the distance between the point of interest and the point detected by the search means on the selected generic model. All these distances form the transformation vector that should be applied to the point of interest to correspond to the detected point. The search means 22 aim to determine the mean of these transformation vectors in order to obtain the overall transformation of the selected generic model.
[0076] In other words, by considering the new transformation vector CompVec of M components, the three-dimensional configuration Pts3D of the vertices can be understood by applying the following equation:
[0077] Pts3D=ModAv+CompVec*ModSigma
[0078] For an unknown configuration Pts3D, e.g. in the case of a new patient, the goal is to find the values of the morphological components CompVec corresponding to a person knowing the average model ModAverage and the transformation matrix ModSigma.
[0079] To this end, the search means 22 calculate the difference DiffMod between the three-dimensional configuration Pts3D of the vertices and the pseudo-inverse matrix ModSigmaInv of the mean models ModMoyen and ModSigma.
[0080] For example, the pseudo-inverse matrix ModSigmaInv can be calculated by decomposing the matrix ModSigma into singular values using the following equation:
[0081] ModSigma = VU*;
[0082] ModSigmaInv=VE t U*;
[0083] E t Corresponding to the transposed matrix of E;
[0084] V* is the anti-conjugate matrix of V; and
[0085] U* is the anti-conjugate matrix of U.
[0086] Using these data, the search device 22 calculates the morphological component CompVec using the following equation:
[0087] DiffMod*ModSigmaInv=CompVec*ModSigma*ModSigmaInv
[0088] That is, CompVec = DiffMod*ModSigmaInv, which also makes it possible to obtain the morphological component CompVec of a specific patient.
[0089] The transformation vector CompVec is then applied to the selected generic model. The pose is estimated again as before, the generic model is adjusted if necessary, and a new search is performed until the generic model is close enough to the 3D representation Re. Finally, the loop stops when the average Euclidean distance between all vertices of the generic model and the vertices corresponding to them on the point cloud is less than a threshold defined according to the number of vertices of the generic model and the size of the object / body of interest (e.g. 2 mm), or when a maximum number of iterations (e.g. 100 iterations) has been reached without reaching an average distance less than the threshold.
[0090] The calibration phase between the sensor 14 and the robot must be performed frequently. To calibrate the visual sensor 14 and the robot, the coordinates of at least three common points of the two markings can be recorded. In practice, the number of points used N is preferably greater than 3. The robot moves over the work area and stops N times.
[0091] At each stop, the position of the robot is recorded by calculating the movements performed by the movement commands of the robot, and the position of the stop is known in three dimensions through detection by the vision sensor 14 .
[0092] At the end of these N stops, the coordinates of the N points are known in both reference frames. The barycenter of the distribution of the N points in the two reference frames is determined using the following equation:
[0093] BarycentreA = 1 / N sum(PA(i)) for i = 1 to N, where PA(i) is a point in the reference frame of the sensor 14; and
[0094] For i=1 to N, BarycentreB=1 / N sum(PB(i)), where PB(i) is a point in the robot reference frame.
[0095] Then, the covariance matrix C is determined by the following equation:
[0096] For i=1 to N, C=sum((PA(i)-BarycentreA)(PB(i)-BarycentreB) t )
[0097] Then decompose the covariance matrix C into its singular values:
[0098] C=UEV*
[0099] The rotation matrix R between the two reference markers is then obtained by the following equation:
[0100] R=VU t ; If the determinant of R is negative, then the third column of the rotation matrix R can be multiplied by -1.
[0101] The translation to be applied between two markers is determined by the following equation:
[0102] T=-R*BarycentreA+BarycentreB
[0103] Therefore, any point of the reference frame Pa of the sensor 14 can be transformed to the reference frame Pb of the robot by applying the following equation:
[0104] Pb=R*Pa+T
[0105] In the first embodiment of the present invention, Figure 2 As shown, the adjustment of the selected generic model makes it possible to acquire the position Pr of the reference element 45 directly on the adjusted generic model.
[0106] In step 40, acquisition is performed by determining the position Pr of a reference element 45 on the three-dimensional representation Re of the surface. The reference element may correspond to an actuator or to a reference point or a set of reference points of a physical element. For example, the reference element may correspond to a glove or a set of points representing the hand of a practitioner. The position Pr of the reference element 45 on the three-dimensional representation Re of the surface may be determined by a position triangulation module or by an image processing analysis similar to that used to capture the three-dimensional representation Re of the surface. In addition to the position Pr of the reference element 45 on the three-dimensional representation Re of the surface, the acquisition may also capture the orientation of the reference element 45 or the action (e.g. heating) performed using the reference element 45, or a specific movement.
[0107] In step 41, this acquisition is repeated a number of times to form a series of records Tr describing the actual movement performed by the reference element 45. For example, this acquisition may be performed every 0.1 s.
[0108] When the motion sequence Tr is completed, the points of the sequence Tr are projected onto the generic model transformed into the morphology of a human. The generic model is then transformed again to recover these initial parameters.
[0109] To this end, a transformation of the generic model is calculated from the initial parameters to the parameters of the transformed generic model.The motion sequence Tr is then transformed using the same transformation as that applied to the generic model.
[0110] Therefore, the actual motion sequence of Tr is transformed into a generic motion sequence Tx associated with the generic model.
[0111] In a second embodiment of the present invention, Figure 3 As shown, acquisition 45 is performed independently of the general model. In this embodiment, Figure 1 Step 23 of only determines the transformations of the generic model without actually applying them. To match the motion sequence Tx to the generic model, the motion sequence may pass through characteristic points of the surface morphology. For example, acquisition 45 may be performed by moving reference elements at the top of the skull and at the armpits of the subject.
[0112] In step 46, when the motion sequence Tr is recorded, step 47 transforms the actual motion sequence Tr into the generic motion sequence Tr by applying the difference between the generic model and the three-dimensional representation Re of the surface. In a final step 48, the generic motion sequence is repositioned on the chosen generic model. Preferably, this step 48 is performed by seeking to match the characteristic points passed by the reference element 45 and the position of these characteristic points on the generic model. This step 48 can also be performed by taking into account the position of the object.
[0113] The invention thus enables the definition of generic motion sequences Tx on a generic model in a practical way, that is to say without the operator having to use the screen of a computer or digital tablet. The invention thus makes it possible to greatly simplify the process of defining motion sequences, since the operator is generally more efficient during actual recording in real situations.
[0114] This motion sequence can then be used for various applications, such as comparing several motion sequences or the control of a robot that includes means for adapting a generic motion sequence to a specific object.
Claims
1. A device for defining a generic motion sequence (Tx) on a generic model (m1-m3), It is characterized in that The device comprises: - acquisition means for acquiring the position (Pr) of a reference element (45) moving on a surface (40, 45); said reference element (45) being configured to execute the actual movement sequence (Tr); - recording means for recording (41, 46) said actual movement sequence (Tr); - means for acquiring (14) a three-dimensional representation (Re) of said surface; - adjustment means for adjusting (23) said generic model (m1-m3) to adapt said three-dimensional representation (Re) of said surface; and - definition means for defining (42, 47) a generic movement sequence (Tx) on said generic model (m1-m3) by applying said adjustment between said generic model (m1-m3) and said three-dimensional representation (Re) of said surface to said actual movement sequence (Tr).
2. The device according to claim 1, It is characterized in that The adjustment device (23) is configured to fit the general model (m1-m3) to the three-dimensional representation (Re) of the surface; the recording device (41) is configured to record the actual motion sequence (Tr) on the general model (m1-m3) when the general model is fitted to the dimensions of the three-dimensional representation (Re) of the surface; and the definition device (42) is configured to transform the general model (m1-m3) using the recorded motion sequence so that the general model restores these initial parameters.
3. The device according to claim 1, It is characterized in that The adjustment means (23) are configured to calculate the difference between the generic model (m1-m3) and the three-dimensional representation (Re) of the surface; the recording means (46) are configured to record the actual motion sequence (Tr) independently of the generic model (m1-m2); the definition means (47) are configured to transform the recorded motion sequence according to the difference calculated by the adjustment means (23); and the device comprises means for positioning (48) the generic motion sequence on the generic model.
4. The device according to claim 2 or 3, It is characterized in that The acquisition means (40, 45) of the position (Pr) of the reference element (45) are configured to detect the orientation of the reference element (45) in order to report this orientation at various points of the generic motion sequence (Tx).
5. The device according to claim 2 or 3, It is characterized in that The acquisition means (40, 45) of the position (Pr) of the reference element (45) are configured to detect the actions performed or the stresses to which the reference element (45) is subjected in order to transmit these actions or these stresses to the various points of the general movement sequence (Tx).
6. The device according to claim 1, It is characterized in that The generic model (m1-m3) and the three-dimensional representation (Re) of the surface are formatted in the shape of a point cloud, and the adjustment means (23) comprises: - means for calculating (19) the direction of the normal to each point of said three-dimensional representation (Re) of said surface; and - means for searching (20) in a neighbourhood, for each point of said cloud of points of said three-dimensional representation (Re), a point of said generic model (m1-m3) for which the difference between the direction of its normal and the direction of the normal of the point of interest is minimal; - means for determining (21) the distance between a detection point of said generic model (m1-m3) and said point of interest; and - Search means for searching (22) for a global transformation of said generic model (m1-m3) as a function of said distances determined for all points of said point cloud of said three-dimensional representation (Re).
7. The device according to claim 6, It is characterized in that The search device (22) is configured to search for the point of the generic model in a preset sphere surrounding the point of interest.
8. The device according to claim 6 or 7, It is characterized in that The normal direction is determined by constructing a face using the coordinates of the three or four points closest to the point of interest.
9. The device according to claim 1, It is characterized in that The adjusting device (23) comprises: - means for detecting (16) characteristic points (Pref) on said three-dimensional representation (Re); as well as - means for transforming (18) said generic model (m1-m3) in rotation and / or translation so that said positions of said characteristic points correspond to the positions of characteristic points of said generic model.
10. The device according to claim 1, It is characterized in that The reference element (45) corresponds to a glove and the movement sequence (Tr, Tx) corresponds to the movements performed by the glove during a massage.
