CN114812408B - Method and system for measuring height of stone sweeper from rail surface - Google Patents
Method and system for measuring height of stone sweeper from rail surface Download PDFInfo
- Publication number
- CN114812408B CN114812408B CN202210360310.9A CN202210360310A CN114812408B CN 114812408 B CN114812408 B CN 114812408B CN 202210360310 A CN202210360310 A CN 202210360310A CN 114812408 B CN114812408 B CN 114812408B
- Authority
- CN
- China
- Prior art keywords
- point cloud
- cloud data
- rail surface
- stone sweeper
- sweeper
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01B—MEASURING LENGTH, THICKNESS OR SIMILAR LINEAR DIMENSIONS; MEASURING ANGLES; MEASURING AREAS; MEASURING IRREGULARITIES OF SURFACES OR CONTOURS
- G01B11/00—Measuring arrangements characterised by the use of optical techniques
- G01B11/02—Measuring arrangements characterised by the use of optical techniques for measuring length, width or thickness
- G01B11/06—Measuring arrangements characterised by the use of optical techniques for measuring length, width or thickness for measuring thickness ; e.g. of sheet material
- G01B11/0608—Height gauges
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T17/00—Three dimensional [3D] modelling, e.g. data description of 3D objects
- G06T17/30—Polynomial surface description
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/70—Determining position or orientation of objects or cameras
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/80—Analysis of captured images to determine intrinsic or extrinsic camera parameters, i.e. camera calibration
- G06T7/85—Stereo camera calibration
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10028—Range image; Depth image; 3D point clouds
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30244—Camera pose
Abstract
The invention discloses a method and a system for measuring the height of a stone sweeper from a rail surface, wherein the measuring method comprises the following steps: and a point cloud data acquisition step: the point cloud data of the stone sweeper and the rail surface are acquired and output through the acquisition unit; the point cloud model obtaining step: dividing the point cloud data into stone sweeper point cloud data and rail surface point cloud data, and performing surface fitting on the stone sweeper point cloud data and the rail surface point cloud data to obtain a point cloud model; and (3) processing a point cloud model: filling the holes among the point cloud models, and then slicing to obtain a plurality of groups of slice data; and a height value calculating step: and obtaining the height value of the stone sweeper from the rail surface according to the intersection coordinates of the rail surface edge line and the stone sweeper edge line and the point cloud data after obtaining the rail surface edge line and the stone sweeper edge line according to the plurality of groups of slice data and the point cloud data. According to the invention, three-dimensional data of the stone sweeper are acquired in a non-contact mode, so that accurate measurement of the height of the stone sweeper from the rail surface is realized.
Description
Technical Field
The invention relates to the technical field of intelligent inspection of railway trains, in particular to a method and a system for measuring the height of a stone sweeper from a rail surface.
Background
In the daily maintenance work of rail vehicles, the measurement of the dimensions of the wearing parts is an important detection item, which directly affects the driving safety. Therefore, the measurement data must meet the requirements of the inspection standard and be accurate. In the traditional maintenance process, the height measurement of the stone sweeper from the rail surface is mainly based on manual visual ruler reading, and the problems of certain deviation of measurement results are easily caused exist, and the specific conditions are as follows:
1. the surface of the bottom of the stone sweeper is uneven and has a certain radian;
2. the narrow working space is inconvenient to use a high-precision measuring tool;
3. the naked eyes cannot guarantee horizontal observation;
4. the operator is easy to fatigue, and the risk of missed detection can exist;
5. the lack of informatization support can not carry out statistical analysis on the historical data.
It is therefore desirable to develop a method and system for measuring the height of a rock sweeper from the rail surface that overcomes the above-mentioned drawbacks.
Disclosure of Invention
The invention provides a method for measuring the height of a stone sweeper from a rail surface, which comprises the following steps:
and a point cloud data acquisition step: the point cloud data of the stone sweeper and the rail surface are acquired and output through the acquisition unit;
the point cloud model obtaining step: dividing the point cloud data into stone sweeper point cloud data and rail surface point cloud data, and performing surface fitting on the stone sweeper point cloud data and the rail surface point cloud data to obtain a point cloud model;
and (3) processing a point cloud model: filling the holes among the point cloud models, and then slicing to obtain a plurality of groups of slice data;
and a height value calculating step: and obtaining a rail surface edge line and a stone sweeper edge line according to the plurality of groups of slice data and the point cloud data, and obtaining a height value of the stone sweeper from the rail surface according to the intersection point coordinates of the rail surface edge line and the stone sweeper edge line and the point cloud data.
The above measurement method, wherein, the collection unit includes a mechanical arm and a 3D camera installed on the mechanical arm, the measurement method further includes:
an autonomous posture correction step: and obtaining a target gesture according to the target gesture and the gesture of the 3D matched current stone sweeper under the 3D camera coordinate system, and adjusting the acquisition unit according to the target gesture.
The measuring method, wherein the step of obtaining the point cloud model includes:
discrete point processing: filtering discrete points in the point cloud data according to constraint conditions set by the point cloud characteristics;
abnormal point processing: filtering the point cloud data with discrete points filtered by a filter to remove abnormal points in the point cloud data;
and a point cloud data segmentation step: dividing the point cloud data into the stone sweeper point cloud data and the rail surface point cloud data through a trained PointNet++ model;
fitting point cloud data: and fitting the point cloud data of the stone sweeper and the rail surface point cloud data by an interpolation type mobile least square method to obtain the point cloud model.
The measuring method, wherein the point cloud model processing step comprises the following steps: and carrying out triangulation processing on the point cloud model obtained by fitting, and then carrying out slicing processing on the point cloud model according to a set interval to obtain a plurality of groups of slice data.
The measuring method, wherein the step of calculating the height value includes:
edge point acquisition: traversing the intersection points of each group of slice data and the point cloud data to obtain edge points;
edge line acquisition: performing linear fitting on the edge points to obtain the rail surface edge line and the stone sweeper edge line;
a height value obtaining step: and translating the rail surface edge line and the stone sweeper edge line in space to obtain two intersection points and intersection point coordinates of the intersection points, and calculating according to the intersection point coordinates to obtain the height value.
The above measurement method, wherein the acquisition unit includes a mechanical arm and a 3D camera mounted on the mechanical arm, and the autonomous posture correction step includes:
the current pose relation obtaining step: 3D matching is carried out on the current point cloud data of the stone sweeper under the postures of the plurality of mechanical arms and the standard point cloud data of the stone sweeper, and then the current posture relation between the 3D camera and the current stone sweeper is obtained;
the target pose obtaining step: calculating the positioning pose between the mechanical arm and the 3D camera according to the current pose relationship and the poses of the plurality of mechanical arms;
a first pose acquisition step: performing 3D point cloud matching according to the current point cloud data of the stone sweeper and the standard point cloud data of the stone sweeper when the mechanical arm is in the initial posture to obtain a first posture of the current stone sweeper under the 3D camera coordinate system;
and a second pose acquisition step: obtaining a second pose of the current stone sweeper under the current mechanical arm base coordinate system according to the first pose, the positioning pose and the initial pose;
and a mechanical arm adjusting step: and obtaining a target gesture according to the second gesture, the standard gesture relation and the calibration gesture, and adjusting the mechanical arm according to the target gesture.
The invention also provides a measuring system for the height of the stone sweeper from the rail surface, which comprises:
the acquisition unit is used for acquiring and outputting point cloud data of the stone sweeper and the rail surface;
the point cloud model obtaining unit is used for dividing the point cloud data into stone sweeper point cloud data and rail surface point cloud data, and performing surface fitting on the stone sweeper point cloud data and the rail surface point cloud data to obtain a point cloud model;
the point cloud model processing unit is used for filling the holes among the point cloud models and then carrying out slicing processing to obtain a plurality of groups of slice data;
and the height value calculating unit is used for obtaining the height value of the stone sweeper from the rail surface according to the intersection coordinates of the rail surface edge line and the stone sweeper edge line and the point cloud data after obtaining the rail surface edge line and the stone sweeper edge line according to the plurality of groups of slice data and the point cloud data.
The measurement system described above, wherein the measurement system includes:
and the autonomous posture correction unit is used for obtaining a target posture according to the target posture and the posture of the current stone sweeper under the 3D camera coordinate system after 3D matching, and adjusting the acquisition unit according to the target posture.
The measurement system described above, wherein the point cloud model obtaining unit includes:
the discrete point processing step module filters discrete points in the point cloud data according to constraint conditions set by the point cloud characteristics;
the abnormal point processing module is used for filtering the point cloud data with discrete points filtered through a filter to remove abnormal points in the point cloud data;
the point cloud data segmentation module is used for segmenting the point cloud data into the stone sweeper point cloud data and the rail surface point cloud data through a trained PointNet++ model;
and the point cloud data fitting module is used for fitting the point cloud data of the stone sweeper and the rail surface point cloud data by an interpolation type mobile least square method to obtain the point cloud model.
The measurement system described above, wherein the height value calculation unit includes:
the edge point acquisition module is used for traversing the intersection points of each group of slice data and the point cloud data to acquire edge points;
the edge line acquisition module is used for carrying out linear fitting on the edge points to obtain the rail surface edge line and the stone sweeper edge line;
and the height value obtaining module is used for translating the rail surface edge line and the stone sweeper edge line in space to obtain two intersection points and intersection point coordinates of the intersection points, and calculating according to the intersection point coordinates to obtain the height value.
Compared with the prior art, the invention has the following effects: the invention applies a robot and a 3D visual detection technology to analyze data of the abrasion condition of the stone sweeper. Three-dimensional data of the stone sweeper can be obtained in a non-contact mode, and accurate measurement of the height of the stone sweeper from the rail surface is achieved by a series of technical means such as point cloud filtering, point cloud matching, hand-eye calibration, surface fitting and slicing, and error control is achieved within +/-1 mm. The device can replace the manual work to accomplish measurement work, and supplementary manual work accomplishes the one-level and repair the operation, reduces working strength, improves maintenance quality and efficiency. In addition, through the information management platform, the measurement result is recorded in an image and data mode, and support is provided for subsequent full life cycle big data analysis of the component.
Additional features and advantages of the invention will be set forth in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. The objectives and other advantages of the invention may be realized and attained by the structure particularly pointed out in the written description and claims hereof as well as the appended drawings.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions of the prior art, the following description will briefly explain the drawings used in the embodiments or the description of the prior art, and it is obvious that the drawings in the following description are some embodiments of the present invention, and other drawings can be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a flow chart of the measurement method of the present invention;
FIG. 2 is a partial flow chart of step S1 in FIG. 1;
FIG. 3 is a partial flow chart of step S3 in FIG. 1;
FIG. 4 is a partial flow chart of step S5 in FIG. 1;
FIG. 5 is a schematic diagram of the structure of the measuring system of the present invention;
fig. 6 is a schematic structural diagram of the acquisition unit of the present invention.
Detailed Description
For the purpose of making the objects, technical solutions and advantages of the embodiments of the present invention more apparent, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention, and it is apparent that the described embodiments are some embodiments of the present invention, but not all embodiments of the present invention. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
The exemplary embodiments of the present invention and the descriptions thereof are intended to illustrate the present invention, but not to limit the present invention. In addition, the same or similar reference numerals are used for the same or similar parts in the drawings and the embodiments.
The terms "first," "second," "S1," "S2," …, and the like, as used herein, do not denote a particular order or sequence, nor are they intended to limit the invention, but rather are merely intended to distinguish one element or operation from another in the same technical terms.
As used herein, the terms "comprising," "including," "having," "containing," and the like are intended to be inclusive and mean an inclusion, but not limited to.
Reference herein to "a plurality" includes "two" and "more than two"; the term "plurality of sets" as used herein includes "two sets" and "more than two sets".
Referring to fig. 1, fig. 1 is a flow chart of the measurement method of the present invention. As shown in fig. 1, the method for measuring the height of the stone sweeper from the rail surface comprises the following steps:
autonomous posture correction step S1: obtaining a target gesture according to the target gesture and the gesture of the 3D matched current stone sweeper under a 3D camera coordinate system, and adjusting an acquisition unit according to the target gesture; the position and the posture of the stone sweeper, which are output after the hand and eye calibration and the 3D matching, under a camera coordinate system are combined, the position and the posture to be adjusted can be calculated and fed back to the mechanical arm, so that the deviation caused by the measurement due to the change of the relative position relationship between the stone sweeper and the 3D camera caused by wheel pair abrasion or chassis positioning error is corrected;
and (2) a point cloud data acquisition step S2: the point cloud data of the stone sweeper and the rail surface are acquired and output through the acquisition unit;
step S3 of obtaining a point cloud model: dividing the point cloud data into stone sweeper point cloud data and rail surface point cloud data, and performing surface fitting on the stone sweeper point cloud data and the rail surface point cloud data to obtain a point cloud model;
and (S4) a point cloud model processing step: filling the holes among the point cloud models, and then slicing to obtain a plurality of groups of slice data; specifically, in this step, triangulating the point cloud models obtained by fitting to fill voids between the point cloud models, and slicing the point cloud models according to a normal vector direction at a set interval to obtain multiple sets of slice data, where in this embodiment, it is guaranteed that more than 20 sets of intersections between slices and the upper surface of the point cloud models are preferred embodiments;
height value calculation step S5: and obtaining a rail surface edge line and a stone sweeper edge line according to the plurality of groups of slice data and the point cloud data, and obtaining a height value of the stone sweeper from the rail surface according to the intersection point coordinates of the rail surface edge line and the stone sweeper edge line and the point cloud data.
Referring to fig. 2, fig. 2 is a flowchart illustrating a sub-step of step S1 in fig. 1. As shown in fig. 2, the autonomous posture correction step S1 includes:
the current pose relation obtaining step S11: 3D matching is carried out on the current point cloud data of the stone sweeper under the postures of the plurality of mechanical arms and the standard point cloud data of the stone sweeper, and then the current posture relation between the 3D camera and the current stone sweeper is obtained;
target pose acquisition step S12: calculating the positioning pose between the mechanical arm and the 3D camera according to the current pose relationship and the poses of the plurality of mechanical arms;
a first pose acquisition step S13: performing 3D point cloud matching according to the current point cloud data of the stone sweeper and the standard point cloud data of the stone sweeper when the mechanical arm is in the initial posture to obtain a first posture of the current stone sweeper under the 3D camera coordinate system;
second pose acquisition step S14: obtaining a second pose of the current stone sweeper under the current mechanical arm base coordinate system according to the first pose, the positioning pose and the initial pose;
mechanical arm adjusting step S15: and obtaining a target gesture according to the second gesture, the standard gesture relation and the calibration gesture, and adjusting the mechanical arm according to the target gesture. The standard pose relationship is a pose relationship in a set ideal state.
The autonomous posture correction step S1 is explained below in connection with the specific embodiment:
1) The mechanical arm is controlled to complete path planning, the initial posture A of the mechanical arm is recorded, and the standard stone sweeper is acquired to obtain point cloud data of the standard stone sweeper;
2) The mechanical arm is adjusted for multiple times, the posture of the mechanical arm after adjustment is recorded, meanwhile, the current stone sweeper is collected based on the posture of the mechanical arm after adjustment to obtain current stone sweeper point cloud data, 3D matching is conducted on the current stone sweeper point cloud data and standard stone sweeper point cloud data, the current posture relation between the 3D camera and the current stone sweeper is obtained, and the calibration posture between the mechanical arm and the 3D camera is calculated according to the current posture relation and the postures of the mechanical arm;
4) Solving a positioning pose P between the mechanical arm and the 3D camera, namely hand-eye calibration, by utilizing SVD decomposition and a nonlinear least square method according to the current pose relation and the multiple mechanical arm poses;
5) The mechanical arm is controlled to automatically move from the initial posture to the initial posture A, a signal is sent to the position, and the 3D camera is triggered to acquire the point cloud data of the current stone sweeper;
6) Performing 3D point cloud matching on the current point cloud data of the stone sweeper and the standard point cloud data of the stone sweeper to obtain a first pose B of the current stone sweeper under the 3D camera coordinate system;
7) Calculating a second pose T of the stone sweeper under the current mechanical arm base coordinate system by combining the standard pose P, the initial pose A and the first pose B, wherein T=AXPXB;
8) And the pose T of the stone sweeper relative to the mechanical arm base is fixed, a target pose D is solved according to a formula T=CXPXD, the pose of the mechanical arm is adjusted through the target pose D, and C is the pose relation between the 3D camera and the stone sweeper in an ideal state.
In this embodiment, the point cloud registration includes two steps of coarse registration and fine registration, wherein the accurate selection of the matching point pairs in the coarse registration directly affects the robustness and accuracy of the fine registration. The coarse registration process is as follows:
1) And calculating FPFH feature descriptors of all points in the source point cloud so as to define a normal distribution relation between the point P and a neighborhood point P', wherein the specific formula is as follows:
in the above, SPFH (P) is a triplet between P and PCharacteristic histogram of composition, w k As the weight, the distance from the point P to its neighborhood point P' is generally used as the weight.
2) And (3) according to the FPFH feature descriptors of each point in the point cloud obtained in the step (1), establishing indexes for feature vectors of the point cloud by assisting with a k-d tree according to a feature similarity principle. If the nearest neighbor feature vector F (Q) is found in the target point cloud Q, the feature vector F (P) of each point in the source point cloud P is regarded as P and Q as a group of matching point pairs, and an initial matching point pair set C is added 1 Is a kind of medium.
3) Because the obstacle deflector and the rail surface are rigid structures, the geometric structure is kept unchanged in the point cloud registration process, and the initial matching point pair set C can be matched based on the characteristics 1 And (5) further screening. Two sets of matching point pairs (p) are found from the target point cloud and the source point cloud 1 ,q 1 ) And (p) 2 ,q 2 ) According to two points p 1 And p 2 And its normal n 1 And n 2 A four-element group (theta) 1 ,θ 2 ,θ n D) accurately describing the geometric structure characteristics of the point cloud area. Wherein θ 1 Is a representation of the normal n 1 Vector of ANDThe included angle theta 2 Is the normal n 2 Vector->Included angle theta n Is the normal n 1 And n 2 Included angle d is p 1 And p 2 Euclidean distance between them.
4)θ n In description n 1 And n 2 Is not unique in terms of spatial positional relationship. To solve the problem, a singularity elimination method based on an indication vector is designed, wherein the indication vector is a normal n 1 Vector of ANDIs a cross product of (a). Optimized θ n The description formula is as follows:
5) Converting dimensionality into dimensionality by using a normalized similarity measurement method, and calculating the similarity of the quaternion of the initial matching point pair according to the following formula:
wherein s is similarity, when s approaches 0, the more similar the geometric structures of two pairs of matching point pairs are, the more accurate the matching point pairs are judged by setting a threshold value, and a new matching point pair set C is formed after screening 2 。
6) At matching point pair set C 2 In the method, a nonlinear optimization method is adopted to solve the transformation matrix. The L-M algorithm is selected, more accurate increment is calculated by introducing a non-singular matrix, and a residual error function is established, wherein the formula is as follows:
in the formula, T is an Euclidean transformation matrix between a source point cloud P and a target point cloud Q, and P and Q are a group of matching point pairs. And the formula for solving the delta equation is as follows:
(J T J+μI)ξ＝-J T f(T)；
wherein J is a derivative matrix of the residual function f (T) with respect to T; i is an identity matrix introduced for solving a singular matrix; mu is a damping coefficient; ζ is the translation vector and euler angle for each iteration. If μ is large, the L-M algorithm approaches the steepest descent method. Conversely, when μ is small, the L-M algorithm approaches Gauss Newton's method. In order to determine the damping coefficient mu, a measurement index rho is set to determine the value of mu, and the specific formula is as follows:
in the formula, if rho approaches 1, the difference between the actual decline value of the function and the decline value estimated by the model is not large; if rho is larger, the drop value and the estimated drop value of the model are larger, and mu is increased in iteration; whereas μ is reduced in iterations. The calculation formula of the transformation matrix T in the L-M algorithm is as follows:
after performing the above steps 1-6, the transformation matrix T after two registrations can be solved. Through the RMSE quantization analysis, the smaller the Euclidean average distance between the matching point pairs is, the higher the registration accuracy is. The formula is as follows:
referring to fig. 3, fig. 3 is a flowchart illustrating a sub-step of step S3 in fig. 1. As shown in fig. 3, the point cloud model obtaining step S3 includes:
discrete point processing step S31: and filtering discrete points in the point cloud data according to constraint conditions set by the point cloud characteristics.
Specifically, because of the influence of external factors such as the reflection of light on the surface of the steel rail and the uneven bottom surface of the obstacle deflector, sparse outlier noise exists in the output point cloud. And setting constraint conditions according to the point cloud characteristics, and filtering out part of discrete points.
Abnormal point processing step S32: and filtering the point cloud data with discrete points filtered by a filter to remove abnormal points in the point cloud data.
Specifically, in the abnormal point processing step S32, downsampling is implemented by the pixel Grid filter under the condition that the geometry of the point cloud data is not affected, so that the efficiency of subsequent point cloud data processing is improved.
In addition to applying the above-mentioned filters, the point V is obtained by constructing the topology of the point cloud data set V i E V, traversal V i Calculating average distance of k neighboring points aroundSimilarly, a distance set is obtained, the average value mu and the standard deviation sigma of the set are obtained, and a distance threshold d is set mid =μ+ωσ, where ω is a weight coefficient. If exceed d mid If the abnormal point is not exceeded, the abnormal point is removed, the point cloud filtering processing is completed mid If so, the abnormal point processing step S32 is continued.
The point cloud data segmentation step S33: dividing the point cloud data into the stone sweeper point cloud data and the rail surface point cloud data through a trained PointNet++ model;
the point cloud data fitting step S34: and fitting the point cloud data of the stone sweeper and the rail surface point cloud data by an interpolation type mobile least square method to obtain the point cloud model.
Specifically, in the point cloud data fitting step S34, the interpolation type moving least square method is used in the surface fitting calculation process, and is applied to the shape function forming the gridless method, and the boundary condition can be directly applied. The moving least squares fit plane formula with interpolation conditions can be expressed as:
wherein, the liquid crystal display device comprises a liquid crystal display device,the construction function is a modified version of the moving least squares fitting formula, the modified term is +.>The specific calculation process is as follows:
1) Calculating a moving least square fitting curve without interpolation conditions;
2) Calculating the deviation delta under the interpolation node s ＝f(x s )-y s ；
3) Calculation of l s (x) And carrying out a motion least square fitting curve under interpolation conditions.
Referring to fig. 4, fig. 4 is a flowchart illustrating a sub-step of step S3 in fig. 1. As shown in fig. 4, the height value calculating step S5 includes:
edge point acquisition step S51: traversing the intersection points of each group of slice data and the point cloud data to obtain edge points;
edge line acquisition step S52: performing linear fitting on the edge points to obtain the rail surface edge line and the stone sweeper edge line;
height value obtaining step S53: and translating the rail surface edge line and the stone sweeper edge line in space to obtain two intersection points and intersection point coordinates of the intersection points, and calculating according to the intersection point coordinates to obtain the height value.
Specifically, a fitting line of the intersection is made to obtain two edge lines of the track surface and the bottom surface of the obstacle deflector, which intersect in translation in space, and intersect at the point a (x 1 ,y 1 ,z 1 ) And point B (x) 2 ,y 2 ,z 2 ) The height value of the obstacle deflector from the rail surface is
Referring to fig. 5-6, fig. 5 is a schematic structural diagram of the measurement system according to the present invention; fig. 6 is a schematic structural diagram of the acquisition unit of the present invention. As shown in fig. 5 to 6, the system for measuring the height of the stone sweeper from the rail surface according to the present invention comprises: the system comprises an autonomous posture correction unit 11, an acquisition unit 12, a point cloud model acquisition unit 13, a point cloud model processing unit 14 and a height value calculation unit 15; the autonomous posture correction unit 11 obtains a target posture according to the target posture and the posture of the 3D matched current stone sweeper under the 3D camera coordinate system, and adjusts the acquisition unit according to the target posture; the acquisition unit 12 acquires and outputs point cloud data of the stone sweeper and the rail surface; the point cloud model obtaining unit 13 divides the point cloud data into stone sweeper point cloud data and rail surface point cloud data, and performs surface fitting on the stone sweeper point cloud data and the rail surface point cloud data to obtain a point cloud model; the point cloud model processing unit 14 fills the holes among the point cloud models and then performs slicing processing to obtain a plurality of groups of slice data; and the height value calculating unit 15 obtains the height value of the stone sweeper from the rail surface according to the intersection coordinates of the rail surface edge line and the stone sweeper edge line and the point cloud data after obtaining the rail surface edge line and the stone sweeper edge line according to the plurality of groups of slice data and the point cloud data.
Wherein the point cloud model processing unit 14 further comprises: and carrying out triangulation processing on the point cloud model obtained by fitting, and then carrying out slicing processing on the point cloud model according to a set interval to obtain a plurality of groups of slice data.
Further, the acquisition unit 12 includes a mechanical arm 121 and a 3D camera 122,3D camera 122 mounted on the mechanical arm 121 through an adapter plate 123. The autonomous posture correction unit 11 includes: the target pose acquisition module 111, the first pose acquisition module 112, the second pose acquisition module 113 and the mechanical arm adjustment module 114; the target pose acquiring module 111 calculates a pose relation between the 3D camera and the standard stone sweeper and a target pose between the mechanical arm and the 3D camera after performing 3D matching according to the point cloud data of the stone sweeper and the point cloud data of the standard stone sweeper under different mechanical arm poses; the first pose acquisition module 112 is used for performing 3D point cloud matching according to the current point cloud data of the stone sweeper when the mechanical arm is in the initial pose and the standard point cloud data of the stone sweeper to acquire a first pose of the current stone sweeper under the 3D camera coordinate system; the second pose acquisition module 113 acquires a second pose of the current stone sweeper under the current mechanical arm base coordinate system according to the first pose, the calibration pose and the initial pose; the mechanical arm adjusting module 114 obtains a target pose according to the second pose, the pose relationship, and the positioning pose, and adjusts the mechanical arm according to the target pose.
Still further, the point cloud model obtaining unit 13 includes:
a discrete point processing step module 131, configured to filter out discrete points in the point cloud data according to constraint conditions set by the point cloud characteristics;
the abnormal point processing module 132 is used for filtering the point cloud data with discrete points filtered by a filter to remove abnormal points in the point cloud data;
the point cloud data segmentation module 133 is used for segmenting the point cloud data into the stone sweeper point cloud data and the rail surface point cloud data through the trained PointNet++ model;
the point cloud data fitting module 134 fits the point cloud data of the stone sweeper and the point cloud data of the rail surface by an interpolation type moving least square method to obtain the point cloud model.
Still further, the height value calculation unit 15 includes:
the edge point obtaining module 151 traverses each set of intersection points of the slice data and the point cloud data to obtain edge points;
the edge line obtaining module 152 is used for performing linear fitting on the edge points to obtain the rail surface edge line and the stone sweeper edge line;
and the height value obtaining module 153 translates the rail surface edge line and the stone sweeper edge line in space to obtain two intersection points and intersection point coordinates of the intersection points, and calculates the height value according to the intersection point coordinates.
In conclusion, the measurement accuracy of the invention reaches 0.1mm, and the measurement error is controlled to be +/-0.5 mm. Compared with manual work, the measurement accuracy is improved to a certain extent. The accurate measurement data has a vital significance for the application abrasion analysis of the whole life cycle of the component; meanwhile, the measurement operation of the full-automatic process can replace manual work, so that the labor intensity is reduced, and the labor efficiency is improved.
Although the invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical scheme described in the foregoing embodiments can be modified or some technical features thereof can be replaced by equivalents; such modifications and substitutions do not depart from the spirit and scope of the technical solutions of the embodiments of the present invention.
Claims (7)
1. A method for measuring the height of a stone sweeper from a rail surface, comprising the steps of:
an autonomous posture correction step: obtaining a target gesture according to a target gesture and a gesture of a 3D matched current stone sweeper under a 3D camera coordinate system, and adjusting an acquisition unit according to the target gesture, wherein the acquisition unit comprises a mechanical arm and a 3D camera arranged on the mechanical arm, and the autonomous gesture correction step comprises a current gesture relation acquisition step, a target gesture acquisition step, a first gesture acquisition step, a second gesture acquisition step and a mechanical arm adjustment step:
the current pose relation obtaining step: 3D matching is carried out on the current point cloud data of the stone sweeper under the postures of the plurality of mechanical arms and the standard point cloud data of the stone sweeper, and then the current posture relation between the 3D camera and the current stone sweeper is obtained;
the target pose obtaining step: calculating the positioning pose between the mechanical arm and the 3D camera according to the current pose relationship and the poses of the plurality of mechanical arms;
a first pose acquisition step: performing 3D point cloud matching according to the current point cloud data of the stone sweeper and the standard point cloud data of the stone sweeper when the mechanical arm is in the initial posture to obtain a first posture of the current stone sweeper under the 3D camera coordinate system;
and a second pose acquisition step: obtaining a second pose of the current stone sweeper under the current mechanical arm base coordinate system according to the first pose, the positioning pose and the initial pose;
and a mechanical arm adjusting step: obtaining a target gesture according to the second gesture, the standard gesture relation and the calibration gesture, and adjusting the mechanical arm according to the target gesture;
and a point cloud data acquisition step: the point cloud data of the stone sweeper and the rail surface are acquired and output through the acquisition unit;
the point cloud model obtaining step: dividing the point cloud data into stone sweeper point cloud data and rail surface point cloud data, and performing surface fitting on the stone sweeper point cloud data and the rail surface point cloud data to obtain a point cloud model;
and (3) processing a point cloud model: filling the holes among the point cloud models, and then slicing to obtain a plurality of groups of slice data;
and a height value calculating step: and obtaining a rail surface edge line and a stone sweeper edge line according to the plurality of groups of slice data and the point cloud data, and obtaining a height value of the stone sweeper from the rail surface according to the intersection point coordinates of the rail surface edge line and the stone sweeper edge line and the point cloud data.
2. The measurement method according to claim 1, wherein the point cloud model obtaining step includes:
discrete point processing: filtering discrete points in the point cloud data according to constraint conditions set by the point cloud characteristics;
abnormal point processing: filtering the point cloud data with discrete points filtered by a filter to remove abnormal points in the point cloud data;
and a point cloud data segmentation step: dividing the point cloud data into the stone sweeper point cloud data and the rail surface point cloud data through a trained PointNet++ model;
fitting point cloud data: and fitting the point cloud data of the stone sweeper and the rail surface point cloud data by an interpolation type mobile least square method to obtain the point cloud model.
3. The measurement method of claim 1, wherein the point cloud model processing step further comprises: and carrying out triangulation processing on the point cloud model obtained by fitting, and then carrying out slicing processing on the point cloud model according to a set interval to obtain a plurality of groups of slice data.
4. The measurement method according to claim 1, wherein the height value calculation step includes:
edge point acquisition: traversing the intersection points of each group of slice data and the point cloud data to obtain edge points;
edge line acquisition: performing linear fitting on the edge points to obtain the rail surface edge line and the stone sweeper edge line;
a height value obtaining step: and translating the rail surface edge line and the stone sweeper edge line in space to obtain two intersection points and intersection point coordinates of the intersection points, and calculating according to the intersection point coordinates to obtain the height value.
5. A system for measuring the height of a rock sweeper from a rail surface, characterized in that the measuring method according to claim 1 is applied, said measuring system comprising:
the autonomous attitude correction unit obtains a target attitude according to the target attitude and the attitude of the current stone sweeper under the 3D camera coordinate system after 3D matching, and adjusts the acquisition unit according to the target attitude;
the acquisition unit is used for acquiring and outputting point cloud data of the stone sweeper and the rail surface;
the point cloud model obtaining unit is used for dividing the point cloud data into stone sweeper point cloud data and rail surface point cloud data, and performing surface fitting on the stone sweeper point cloud data and the rail surface point cloud data to obtain a point cloud model;
the point cloud model processing unit is used for filling the holes among the point cloud models and then carrying out slicing processing to obtain a plurality of groups of slice data;
and the height value calculating unit is used for obtaining the height value of the stone sweeper from the rail surface according to the intersection coordinates of the rail surface edge line and the stone sweeper edge line and the point cloud data after obtaining the rail surface edge line and the stone sweeper edge line according to the plurality of groups of slice data and the point cloud data.
6. The measurement system of claim 5, wherein the point cloud model obtaining unit comprises:
the discrete point processing step module filters discrete points in the point cloud data according to constraint conditions set by the point cloud characteristics;
the abnormal point processing module is used for filtering the point cloud data with discrete points filtered through a filter to remove abnormal points in the point cloud data;
the point cloud data segmentation module is used for segmenting the point cloud data into the stone sweeper point cloud data and the rail surface point cloud data through a trained PointNet++ model;
and the point cloud data fitting module is used for fitting the point cloud data of the stone sweeper and the rail surface point cloud data by an interpolation type mobile least square method to obtain the point cloud model.
7. The measurement system according to claim 5, wherein the height value calculation unit includes:
the edge point acquisition module is used for traversing the intersection points of each group of slice data and the point cloud data to acquire edge points;
the edge line acquisition module is used for carrying out linear fitting on the edge points to obtain the rail surface edge line and the stone sweeper edge line;
and the height value obtaining module is used for translating the rail surface edge line and the stone sweeper edge line in space to obtain two intersection points and intersection point coordinates of the intersection points, and calculating according to the intersection point coordinates to obtain the height value.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210360310.9A CN114812408B (en) | 2022-04-07 | 2022-04-07 | Method and system for measuring height of stone sweeper from rail surface |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210360310.9A CN114812408B (en) | 2022-04-07 | 2022-04-07 | Method and system for measuring height of stone sweeper from rail surface |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114812408A CN114812408A (en) | 2022-07-29 |
| CN114812408B true CN114812408B (en) | 2023-08-22 |
Family
ID=82534170
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210360310.9A Active CN114812408B (en) | 2022-04-07 | 2022-04-07 | Method and system for measuring height of stone sweeper from rail surface |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114812408B (en) |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108846888A (en) * | 2018-04-23 | 2018-11-20 | 北京建筑大学 | A kind of Gu wooden building components fine size information automation extracting method |
| CN111272091A (en) * | 2020-03-09 | 2020-06-12 | 东南大学 | Automatic identification method for large deformation of welded I-steel based on three-dimensional laser scanning |
| CN112325796A (en) * | 2020-10-26 | 2021-02-05 | 上海交通大学 | Large-scale workpiece profile measuring method based on auxiliary positioning multi-view point cloud splicing |
| CN112414309A (en) * | 2020-11-25 | 2021-02-26 | 北京交通大学 | High-speed rail contact line height-guiding and pull-out value inspection method based on airborne laser radar |
| CN112476434A (en) * | 2020-11-24 | 2021-03-12 | 新拓三维技术(深圳)有限公司 | Visual 3D pick-and-place method and system based on cooperative robot |
| CN112581457A (en) * | 2020-12-23 | 2021-03-30 | 武汉理工大学 | Pipeline inner surface detection method and device based on three-dimensional point cloud |
| WO2021123463A1 (en) * | 2019-12-17 | 2021-06-24 | Telefonos, Lineas Y Centrales, S.A. | Method for in-situ and real-time collection and processing of geometric parameters of railway lines |
| CN113192116A (en) * | 2021-05-10 | 2021-07-30 | 聚时科技(上海)有限公司 | Aviation blade thickness parameter measuring method based on structured light camera |
| WO2021189468A1 (en) * | 2020-03-27 | 2021-09-30 | 深圳市速腾聚创科技有限公司 | Attitude correction method, apparatus and system for laser radar |
| CN114049331A (en) * | 2021-11-17 | 2022-02-15 | 长春理工大学 | Method for polishing surface of complex workpiece |
-
2022
- 2022-04-07 CN CN202210360310.9A patent/CN114812408B/en active Active
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108846888A (en) * | 2018-04-23 | 2018-11-20 | 北京建筑大学 | A kind of Gu wooden building components fine size information automation extracting method |
| WO2021123463A1 (en) * | 2019-12-17 | 2021-06-24 | Telefonos, Lineas Y Centrales, S.A. | Method for in-situ and real-time collection and processing of geometric parameters of railway lines |
| CN111272091A (en) * | 2020-03-09 | 2020-06-12 | 东南大学 | Automatic identification method for large deformation of welded I-steel based on three-dimensional laser scanning |
| WO2021189468A1 (en) * | 2020-03-27 | 2021-09-30 | 深圳市速腾聚创科技有限公司 | Attitude correction method, apparatus and system for laser radar |
| CN112325796A (en) * | 2020-10-26 | 2021-02-05 | 上海交通大学 | Large-scale workpiece profile measuring method based on auxiliary positioning multi-view point cloud splicing |
| CN112476434A (en) * | 2020-11-24 | 2021-03-12 | 新拓三维技术(深圳)有限公司 | Visual 3D pick-and-place method and system based on cooperative robot |
| CN112414309A (en) * | 2020-11-25 | 2021-02-26 | 北京交通大学 | High-speed rail contact line height-guiding and pull-out value inspection method based on airborne laser radar |
| CN112581457A (en) * | 2020-12-23 | 2021-03-30 | 武汉理工大学 | Pipeline inner surface detection method and device based on three-dimensional point cloud |
| CN113192116A (en) * | 2021-05-10 | 2021-07-30 | 聚时科技(上海)有限公司 | Aviation blade thickness parameter measuring method based on structured light camera |
| CN114049331A (en) * | 2021-11-17 | 2022-02-15 | 长春理工大学 | Method for polishing surface of complex workpiece |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114812408A (en) | 2022-07-29 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110497187B (en) | Sun flower pattern assembly system based on visual guidance | |
| Chen et al. | High-accuracy multi-camera reconstruction enhanced by adaptive point cloud correction algorithm | |
| US11042146B2 (en) | Automated 360-degree dense point object inspection | |
| KR102257610B1 (en) | EXTRINSIC CALIBRATION METHOD OF PLURALITY OF 3D LiDAR SENSORS FOR AUTONOMOUS NAVIGATION SYSTEM | |
| CN111775152B (en) | Method and system for guiding mechanical arm to grab scattered stacked workpieces based on three-dimensional measurement | |
| CN110866969B (en) | Engine blade reconstruction method based on neural network and point cloud registration | |
| EP3963414A2 (en) | Automated 360-degree dense point object inspection | |
| US9482754B2 (en) | Detection apparatus, detection method and manipulator | |
| CN113324478A (en) | Center extraction method of line structured light and three-dimensional measurement method of forge piece | |
| DE112011100652T5 (en) | THREE-DIMENSIONAL MEASURING DEVICE, PROCESSING METHOD AND NON-VOLATILE COMPUTER-READABLE STORAGE MEDIUM | |
| CN114055255B (en) | Large-scale complex component surface polishing path planning method based on real-time point cloud | |
| CN106152935A (en) | It is applicable to the laser measurement system without feature space curved surface blank and processing positioning method thereof | |
| CN105205806A (en) | Machine vision based precision compensation method | |
| CN112263052A (en) | Method and system for automatically mapping vamp glue spraying path based on visual data | |
| CN112109072B (en) | Accurate 6D pose measurement and grabbing method for large sparse feature tray | |
| CN112880562A (en) | Method and system for measuring pose error of tail end of mechanical arm | |
| CN114434036B (en) | Three-dimensional vision system for gantry robot welding of large ship structural member and operation method | |
| CN113267180A (en) | AGV forklift tray positioning and forking method based on 3D depth vision | |
| CN116625258A (en) | Chain spacing measuring system and chain spacing measuring method | |
| CN114812408B (en) | Method and system for measuring height of stone sweeper from rail surface | |
| Lee et al. | Implementation of a robotic arm with 3D vision for shoes glue spraying system | |
| CN116309882A (en) | Tray detection and positioning method and system for unmanned forklift application | |
| González-Aguilera et al. | Accuracy assessment of vehicles surface area measurement by means of statistical methods | |
| CN106705836A (en) | Automatic detection system for size parameters of T type guide rail | |
| JP7323057B2 (en) | Control device, control method, and control program |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |