CN-122015665-A - Pantograph thickness measuring method and device for three-dimensional space calculation and visual reduction

CN122015665ACN 122015665 ACN122015665 ACN 122015665ACN-122015665-A

Abstract

The pantograph thickness measurement method comprises the following steps of deploying a device, calibrating parameters, acquiring images, solving phases, converting coordinates, generating point clouds, splicing the point clouds, analyzing information, outputting and displaying, and outputting and displaying. In addition, the invention also provides a pantograph thickness measuring device for three-dimensional space calculation and visual reduction. The invention acquires the complete three-dimensional point clouds of the upper surface, the lower surface, the side surface and the edge of the pantograph carbon slide plate by adopting a plurality of groups of area array structured light three-dimensional cameras and adopting a phase coding stripe structured light image acquisition mode, finally realizes local defect detection such as notch, peeling and the like, has stronger environment light interference resistance compared with single-line laser, can greatly improve measurement precision and reliability, has strong practicability and has stronger popularization significance.

Inventors

DONG HUI
XU CHANGYUAN
Gan Yushun
JIANG WENXIN
WU DONGLIANG
LIAO LITAO
WU GENGCAI
LIANG YINGCHANG
LI YINGRONG
CHEN RUNSONG
CHEN YANLI
WU ZHENHUA
LIU QI
ZHENG JIEFENG

Assignees

东莞市诺丽科技股份有限公司

Dates

Publication Date: 20260512
Application Date: 20260126

Claims (8)

1.A pantograph thickness measuring method for three-dimensional space calculation and visual reduction is characterized by comprising the following steps: the method comprises the steps that 1, a deployment device is arranged on a train detection shed, a plurality of groups of area array structured light three-dimensional cameras are respectively arranged on the train detection shed, a carbon slide plate is covered in a overlooking view, a upward view and a side view, a trigger is arranged in a detection area, the plurality of groups of area array structured light three-dimensional cameras are controlled to be started or standby automatically, and each area array structured light three-dimensional camera comprises a projection module and a camera module; step 2, parameter calibration, namely respectively carrying out accurate calibration on a plurality of groups of area array structured light three-dimensional cameras by using a multi-dot array calibration plate, and respectively obtaining an internal reference matrix K of a camera module and an external reference matrix [ R|t ] between a projection module and the camera module; Step 3, image acquisition, when a trigger is triggered, a plurality of groups of area array structure light cameras are controlled to be started in a combined mode according to a preset scheduling strategy, a phase coding stripe projection mode is adopted, a group of sine stripe patterns with different phases are projected to the surface of a carbon slide plate by a projection module of the area array structure light cameras, and the camera module correspondingly shoots and acquires image sequences projected on the surface of the carbon slide plate by the stripe patterns; step 4, solving the phase, namely solving a trigonometric function equation set constructed by a plurality of images to obtain a wrapped phase Recovering continuous phase distribution by phase unwrapping algorithm Specifically, the trigonometric function equation set model is Wherein the method comprises the steps of At the pixel for the mth fringe pattern Is used for the gray-scale value of (c), For the brightness of the background, In order to achieve the degree of modulation, For the initial phase of the pixel point, Phase offset for the mth plot; Step 5, coordinate conversion, the phase value of each pixel is converted The corresponding imaging light ray and the projection plane establish a geometric relation, the depth P is calculated according to a structured light triangulation model, and then the corresponding three-dimensional point is obtained through back projection, specifically, the back projection algorithm is that Wherein the method comprises the steps of Is the three-dimensional coordinate under the camera coordinate system, P is the projection depth, Is pixel coordinates; Step 6, generating point cloud, namely respectively completing three-dimensional reconstruction of all pixel points on a plurality of groups of area array structured light three-dimensional cameras, and obtaining a plurality of groups of two-dimensional high-density point cloud images of the carbon slide plate; Step 7, point cloud splicing, namely acquiring a rotation matrix of each group of area array structured light three-dimensional cameras Translation vector Point cloud under any area array structured light three-dimensional camera coordinate system Converting into world coordinate system, in particular converting algorithm into Wherein the method comprises the steps of For the converted global three-dimensional coordinates, For rotation between the projection module and the camera module, Is a translational external parameter between the projection module and the camera module; Analyzing information, namely carrying out equidistant slicing on a point cloud model along the length direction (X axis) of the carbon slide plate based on a geometric profile analysis method, dividing the whole slide plate into a plurality of partial profiles, extracting Y-axis coordinate values (namely height information) of all points in each profile area, and acquiring the minimum Y value of the profile as a partial residual thickness index by combining noise filtering and statistical analysis; And 9, outputting and displaying, namely outputting all the thickness and defect analysis results into a structured data file, generating visual results such as a thickness profile, a notch distribution diagram and the like based on the structured data file, and outputting the visual results to a Web end for displaying.
2. The method for measuring thickness of a pantograph with three-dimensional space calculation and visual restoration as set forth in claim 1, wherein in the step 1, the trigger is a magnetic steel sensor and an infrared correlation sensor, when a train wheel approaches to the magnetic steel sensor, an electric signal is generated as an incoming signal due to magnetic field disturbance, and then only when the magnetic steel sensor and the infrared correlation sensor are triggered synchronously, the planar array structured light three-dimensional camera is controlled to start for image acquisition.
3. The method for measuring thickness of a pantograph with three-dimensional space calculation and visual restoration as set forth in claim 1, wherein in the step 7, after each two-dimensional point cloud image is converted and spliced into a three-dimensional point cloud image, edge fusion, noise filtering and resolution equalization processing are further performed on an overlapping area by adopting a Voxel Grid downsampling (Voxel Grid), weighted overlap fusion or Iterative Closest Point (ICP) fine tuning method, so that consistency and integrity of point cloud are improved.
4. The method for measuring thickness of a pantograph with three-dimensional space calculation and visual reduction as set forth in claim 1, wherein in step 8, besides the extraction of the residual thickness based on the geometric profile analysis method, the left and right edge regions of the point cloud model are scanned, and the possible phenomena of flaking, breakage or notch are automatically detected by combining geometric features such as normal abrupt change, abnormal curvature and abrupt height drop of the point cloud.
5. The method for measuring thickness of a pantograph with three-dimensional space calculation and visual restoration as set forth in claim 1, wherein in the step 2 and the step 7, calibration can be performed by single-phase rotation or line laser scanning for a plurality of times instead, and then point cloud stitching is realized by feature matching or synchronous positioning and map building (SLAM) method registration.
6. The method for measuring thickness of a pantograph with three-dimensional space calculation and visual restoration as set forth in claim 1, wherein in the step 3, image acquisition is performed by using a mode of scanning a plurality of groups of line lasers and cameras, a mode of using a time-of-flight (TOF) camera, a Gray code mode or a binary code mode instead of a phase code stripe projection mode.
7. A pantograph thickness measuring device for three-dimensional space calculation and visual reduction is used for executing the pantograph thickness measuring method for three-dimensional space calculation and visual reduction according to any one of claims 1 to 4 and is characterized by comprising a plurality of groups of planar array structure light three-dimensional cameras, a trigger and a detection analysis system, wherein the groups of planar array structure light three-dimensional cameras are respectively arranged above a detection shed, between the left side and the right side of the detection shed and between a carbon slide plate and the top of a train, the planar array structure light three-dimensional cameras on the upper side are positioned above the central axis of the carbon slide plate, the planar array structure light three-dimensional cameras on the left side and the right side of the carbon slide plate, the planar array structure light three-dimensional cameras on the lower side are positioned in the area below the carbon slide plate, the trigger is arranged on a train track and is respectively connected with the plurality of groups of planar array structure light three-dimensional cameras and the detection analysis system through signals, and the detection analysis system is respectively connected with the groups of planar array structure light three-dimensional cameras through signals. The area array structured light three-dimensional camera comprises a projection module and a camera module.
8. The three-dimensional space calculation and visual reduction pantograph thickness measurement device according to claim 1, wherein the area array structure light three-dimensional cameras on the left side and the right side are symmetrically designed, two groups of area array structure light three-dimensional cameras on one side are arranged horizontally front and back, the front row of area array structure light three-dimensional cameras are arranged at a backward 30-degree strabismus, and the rear row of area array structure light three-dimensional cameras are arranged at a forward 30-degree strabismus.

Description

Pantograph thickness measuring method and device for three-dimensional space calculation and visual reduction Technical Field The invention relates to the technical field of train tracks, in particular to a pantograph thickness measuring method and device for three-dimensional space calculation and visual restoration. Background The carbon sliding plate of the pantograph is a key stress and friction part of the electric locomotive, and the thickness of the carbon sliding plate directly influences the conductivity and the service life of the pantograph, so that the high-precision and non-contact thickness detection is realized, and the carbon sliding plate has important significance for train operation safety and operation and maintenance efficiency. The existing thickness detection of the carbon slide plate of the pantograph mainly depends on means such as manual measurement, a two-dimensional image method, a high-precision three-dimensional scanner and the like. However, with the popularization of high-speed trains and urban rail vehicles, the traditional manual measurement method and the two-dimensional measurement method have difficulty in meeting the requirements of on-site rapidness, accuracy and automation. The traditional manual measurement method needs to conduct static detection under the conditions of power failure of the overhead contact system and train shutdown, so that the efficiency is low, time and labor are consumed, and the detection result often cannot fully cover the thickness condition of the pantograph slide plate, so that the overall accuracy is insufficient. Meanwhile, the mode also relates to overhead operation and electrified risks, and certain potential safety hazards exist. The two-dimensional image method usually adopts an industrial camera to image the carbon slide plate, and only two-dimensional plane information can be acquired, depth information is lacked, and the detection result is easily influenced by an imaging angle and external illumination conditions (such as rainy days or strong light), so that the precision is limited and the stability is poor. And high-precision three-dimensional scanners generally adopt a detection mode of multiple groups of line laser structured lights. For example, in the prior art of "CN 202011567755.1" an on-line detection device for comprehensive geometric parameters of a pantograph of a train "CN 202111227612.0" a device and a method for identifying, snapping and detecting a pantograph based on three-dimensional line scanning, a technical scheme is disclosed in which multiple groups of structural light sensors are arranged on two sides of a detecting area of the pantograph, each group of structural light sensors is composed of a camera and a line laser, and is used for acquiring cross-section point cloud data of upper and lower surfaces of the pantograph, and forming a three-dimensional model of the pantograph by data stitching, so as to realize measurement of abrasion and offset. Although the multi-group line laser structure light scheme can realize three-dimensional point cloud acquisition and abrasion and offset measurement of the upper surface and the lower surface of the pantograph, the method has the defects that firstly, only the upper surface and the lower surface of the pantograph are covered, the side surface and the edge area are difficult to completely acquire, local defects such as gaps or flaking and the like possibly existing in the pantograph cannot be comprehensively detected, secondly, line laser is sensitive to environmental light interference, and conditions such as strong light, reflection or rain and snow possibly influence the acquisition precision of laser stripes, so that the reliability of a measurement result is influenced. Disclosure of Invention Based on this, it is necessary to provide a pantograph thickness measurement method of three-dimensional space calculation and visual reduction in view of the shortcomings in the prior art. A pantograph thickness measuring method for three-dimensional space calculation and visual reduction comprises the following steps: the method comprises the steps that 1, a deployment device is arranged on a train detection shed, a plurality of groups of area array structured light three-dimensional cameras are respectively arranged on the train detection shed, a carbon slide plate is covered in a overlooking view, a upward view and a side view, a trigger is arranged in a detection area, the plurality of groups of area array structured light three-dimensional cameras are controlled to be started or standby automatically, and each area array structured light three-dimensional camera comprises a projection module and a camera module; step 2, parameter calibration, namely respectively carrying out accurate calibration on a plurality of groups of area array structured light three-dimensional cameras by using a multi-dot array calibration plate, and respectively obtaining an internal refere