US-12623357-B2 - Pose determination in parallel kinematics systems with reference markers

US12623357B2US 12623357 B2US12623357 B2US 12623357B2US-12623357-B2

Abstract

A parallel kinematic system comprises mutually distinguishable markings which are attached in a marking region to the parallel kinematic system. The marking region is a region of the kinematic system that moves along with the pose of the kinematic system. The markings can be attached in a direction at a distance that ensures that n markings are always fully visible in the direction, and the pose of the parallel kinematic system can be determined based on an image that is captured by the camera and contains at least n markings in the direction. The markings can be attached in a direction at a distance that ensures that n or more markings are fully visible in the direction, the markings are attached in different planes, and the pose of the parallel kinematic system can be determined based on an image that is captured by the camera and contains at least any n markings in the direction.

Inventors

Stefan Schulz
Constantin Schempp

Assignees

PHYSIK INSTRUMENTE (PI) SE & CO. KG

Dates

Publication Date: 20260512
Application Date: 20221122
Priority Date: 20211126

Claims (10)

1 . An arrangement with a parallel kinematic system and means for determining the pose of the parallel kinematic system comprising: a camera and a marking region with mutually distinguishable markings, wherein the camera is configured to observe the marking region in different poses of the parallel kinematic system, wherein the means for determining the pose of the parallel kinematic system are configured to determine the pose of the parallel kinematic system based only on images of the marking region captured by the camera if one of the images contains at least a number n of any of the markings in a direction, wherein n is greater than or equal to 1, where a distance, D, between any two markings that are adjacent in the direction satisfies the following formula: F ⁢ O ⁢ V min - ( n + 2 ) * t m n + 1 < D ≤ F ⁢ O ⁢ V min - ( n + 1 ) * t m n , wherein t m is the length of one of the markings, FOV min is a length of the section of the marking region which falls into the field of view of the camera at a minimum distance of the camera from the marking region, wherein the minimum distance is a minimum distance among the distances that the marking region can be away from the camera due to pose changes.
2 . The parallel kinematic system according to claim 1 , wherein the length FOV min satisfies the following equation: FOV min = ( g min f - 1 ) ⁢ l S ⁢ e ⁢ n ⁢ s ⁢ o ⁢ r , g min is the minimum distance, l sensor is a length of the sensor of the camera and f is a focal distance of the camera.
3 . The parallel kinematic system according to claim 1 , wherein the markings are arranged in the marking region according to a regular arrangement pattern.
4 . The parallel kinematic system according to claim 1 , wherein the marking region is attached to an underside of a work platform of the parallel kinematic system and the camera is attached in or on a base of the parallel kinematic system and is directed towards the underside of the work platform, or the marking region is attached in or on the base of the parallel kinematic system and the camera is attached to an underside of the work platform and is directed towards the base of the parallel kinematic system.
5 . The parallel kinematic system according to claim 1 , wherein the length t m of a marking satisfies the following equation: t m ≥ p ⁢ x * p * t b * ( g max f - 1 ) , wherein p is a value that is dependent upon the camera greater than or equal to 2 and less than or equal to 5, px is the length that corresponds to a sampling value of the camera, g max is a maximum distance among the distances that the marking region can be away from the camera due to pose changes, f is a focal distance of the camera, and t b is the number of information units of the marking.
6 . The parallel kinematic system according to claim 1 , wherein the markings are reference markings selected from ARToolKit markings, ArUco markings, QR codes, and AprilTag markings.
7 . The parallel kinematic system according to claim 1 , wherein each of the markings consists of several squares, wherein the squares correspond to the information units and a bit can be encoded in each square.
8 . A method for attaching mutually distinguishable markings to a parallel kinematic system of an arrangement according to claim 1 in a marking region so that the pose of the parallel kinematic system can be determined based only on images of the marking region captured by a camera if one of the images contains at least a predetermined number, n, of markings in a direction, wherein nis greater than or equal to 1, the method comprising: determining a distance, D, between any two markings that are adjacent in the direction according to the following formula: F ⁢ O ⁢ V min - ( n + 2 ) * t m n + 1 < D ≤ F ⁢ O ⁢ V min - ( n + 1 ) * t m n , wherein t m is the length of one of the markings, FOV min is a length of the section of the marking region which falls into the field of view of the camera at a minimum distance of the camera from the marking region, wherein the minimum distance is a minimum distance among the distances that the marking region can be away from the camera due to pose changes, and attaching respectively adjacent markings at the determined distance.
9 . An arrangement with a parallel kinematic system and means for determining the pose of the parallel kinematic system comprising: a camera and a marking region with mutually distinguishable markings, wherein the camera is configured to observe the marking region in different poses of the parallel kinematic system, wherein a distance, D, between any two markings that are adjacent in a direction satisfies the following formula: D ≤ F ⁢ O ⁢ V min - ( n + 1 ) * t m n , wherein t m is the length of one of the markings, FOV min is a length of the section of the marking region which falls into the field of view of the camera at a minimum distance of the camera from the marking region, wherein the minimum distance is a minimum distance among the distances that the marking region can be away from the camera due to pose changes, the markings are disposed in different planes, and the means for determining the pose of the parallel kinematic system are configured to determine the pose of the parallel kinematic system based only on images of the marking region captured by the camera if one of the images contains at least a number, n, of any of the markings in a direction, wherein n is greater than or equal to 2.
10 . A method for attaching mutually distinguishable markings to a parallel kinematic system of an arrangement according to claim 9 in a marking region so that the pose of the parallel kinematic system can be determined based only on images of the marking region captured by a camera if one of the images contains at least a predetermined number, n, of markings in a direction, wherein n is greater than or equal to 2, the method comprising: determining a distance, D, between any two markings that are adjacent in the direction according to the following formula: D ≤ F ⁢ O ⁢ V min - ( n + 1 ) * t m n , wherein t m is the length of one of the markings, FOV min is a length of the section of the marking region which falls into the field of view of the camera at a minimum distance of the camera from the marking region, wherein the minimum distance is a minimum distance among the distances that the marking region can be away from the camera due to pose changes, and attaching any markings that are adjacent in the direction at the determined distance, where the markings are attached such that they are disposed in different planes.

Description

CROSS REFERENCE TO RELATED APPLICATIONS This application is a U.S. National Phase application under 35 U.S.C. 371 of International Application No. PCT/EP2022/082765, filed on Nov. 22, 2022, which claims priority to German Patent Application No. 10 2021 213 358.4, filed on Nov. 26, 2021. The entire disclosures of the above applications are expressly incorporated by reference herein. TECHNICAL FIELD The present invention relates to devices and methods for determining the pose of parallel kinematic systems. BACKGROUND ART A common problem when controlling a robot is determining its current pose (position and orientation). However, in particular in parallel kinematic systems, the pose, i.e. the position and orientation of the movable work platform, for example, of a hexapod, cannot be calculated directly and/or precisely from the lengths of the driven legs or joint angles. Instead, numerical methods, e.g. iterative optimization methods, are necessary for this. On the one hand, these methods are time-consuming and computationally intensive, and the level of accuracy that can be obtained depends largely on the initial estimate used. On the other hand, they are typically based on a measurement of the leg lengths and/or joint angles using internal sensors (e.g. incremental sensors for measuring the leg lengths of a hexapod), which means that influences such as offset, deformation, play and counterplay (backlash) in the legs, joints and/or or the movable platform itself cannot be detected (even with absolute sensors in the legs) and therefore cannot be taken into account by a numerical method based thereon. Furthermore, with incremental sensors, for example, complex reference runs are necessary to obtain a zero position of the legs, and there are special parallel kinematic systems for which no numerical methods are available. However, if external, optical incremental and absolute sensors are used, a large number of sensors are typically necessary in order to be able to measure all degrees of freedom. Due to the restricted measuring range of the sensors, it is also often not possible to obtain large adjustment motions of the parallel kinematic system. When using 6D measurement technology by way of photogrammetry, a large number of images from as many different positions as possible are often necessary to generate an exact 3D scan. Since a comparatively large region has to be captured, only a relatively low level of accuracy can be obtained and an extremely high resolution of the camera is necessary in order to be able to detect changes in position, for example, in the nanometer range. Moldagalieva, Akmaral, et al. “Computer vision-based pose estimation of tensegrity robots using fiducial markers.” 2019 IEEE/SICE International Symposium on System Integration (SiI) IEEE, 2019, discloses a method based on fiducial tags that attempts to keep the entire workspace of the robot in view (in the image region) of the camera. This means that the viewing region of the camera must be very large and the camera must therefore be set up quite a distance away from the movable work platform. Since the accuracy obtainable for a given camera resolution reduces as the camera's distance from the tag increases, this method can only be used to determine the pose of the kinematic system with a low level of accuracy. The present invention is therefore based on the object of improving the determination of the pose of parallel kinematic systems. The object is satisfied according to the invention by the features of the independent claims. Some advantageous embodiments are the object of the dependent claims. SUMMARY The invention is based on the idea of attaching several reference markers to a parallel kinematic system such that at least one of the reference markers or a certain minimum number of reference markers is always in the field of view of the camera. According to a first aspect of the present invention, a parallel kinematic system is provided. The parallel kinematic system comprises a camera and a marking region with mutually distinguishable markings, where the camera is configured to observe the marking region in different poses (or even in all possible poses) of the parallel kinematic system. The pose of the parallel kinematic system can be determined based on an image of the marking region captured by the camera if the image contains at least a number n of any of the markings in a direction, where n is greater than or equal to 1. The distance D between any two markings that are adjacent in a direction satisfies the formula F⁢O⁢Vmin-(n+2)*tmn+1<D≤F⁢O⁢Vmin-(n+1)*tmn, where tm is the length of one of the markings, FOVmin is the length of the section of the marking region that falls into the field of view of the camera at a minimum distance of the camera from the marking region, and the minimum distance is the minimum distance among distances that the marking region can be away from the camera due to pose changes. According to a second as