CN113405455A - Method for automatically calibrating pixel ratio by calculating geometric parameters of contact network - Google Patents

Method for automatically calibrating pixel ratio by calculating geometric parameters of contact network Download PDF

Info

Publication number
CN113405455A
CN113405455A CN202110534974.8A CN202110534974A CN113405455A CN 113405455 A CN113405455 A CN 113405455A CN 202110534974 A CN202110534974 A CN 202110534974A CN 113405455 A CN113405455 A CN 113405455A
Authority
CN
China
Prior art keywords
pixel ratio
characteristic points
calculating
value
parameter
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.)
Granted
Application number
CN202110534974.8A
Other languages
Chinese (zh)
Other versions
CN113405455B (en
Inventor
闫石
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Zhongchuang Zhiwei Technology Co ltd
Original Assignee
Zhongchuang Zhiwei Technology Co ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Zhongchuang Zhiwei Technology Co ltd filed Critical Zhongchuang Zhiwei Technology Co ltd
Priority to CN202110534974.8A priority Critical patent/CN113405455B/en
Publication of CN113405455A publication Critical patent/CN113405455A/en
Application granted granted Critical
Publication of CN113405455B publication Critical patent/CN113405455B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • GPHYSICS
    • G01MEASURING; TESTING
    • G01BMEASURING LENGTH, THICKNESS OR SIMILAR LINEAR DIMENSIONS; MEASURING ANGLES; MEASURING AREAS; MEASURING IRREGULARITIES OF SURFACES OR CONTOURS
    • G01B11/00Measuring arrangements characterised by the use of optical techniques
    • GPHYSICS
    • G01MEASURING; TESTING
    • G01BMEASURING LENGTH, THICKNESS OR SIMILAR LINEAR DIMENSIONS; MEASURING ANGLES; MEASURING AREAS; MEASURING IRREGULARITIES OF SURFACES OR CONTOURS
    • G01B11/00Measuring arrangements characterised by the use of optical techniques
    • G01B11/002Measuring arrangements characterised by the use of optical techniques for measuring two or more coordinates

Landscapes

  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Current-Collector Devices For Electrically Propelled Vehicles (AREA)
  • Image Analysis (AREA)

Abstract

The invention discloses a method for calculating geometric parameters of a contact network and automatically calibrating a pixel ratio, which comprises the following steps: identifying the characteristic points, selecting a plurality of fixed characteristic points, identifying the characteristic points through an identification device, and calculating to obtain a pixel ratio according to structural parameters given by a pantograph manufacturer or the actual distance of the manually and accurately measured characteristic points; deep learning training, namely acquiring identification characteristic points aiming at a complex background and different pantograph models in the running process of a vehicle, automatically identifying and training the characteristic points, and constructing an identification model; and automatic parameter correction, namely automatically identifying the characteristic points according to a set period, simultaneously calculating a pixel ratio, performing mean square error correction on the calculated pixel ratio and the previous pixel ratio, and writing the mean square error correction into a system parameter table. The invention solves the problems of low calibration precision and high cost of the geometric parameters of the existing contact network.

Description

Method for automatically calibrating pixel ratio by calculating geometric parameters of contact network
Technical Field
The invention relates to the technical field of railway safety monitoring, in particular to a method for calculating geometric parameters of a contact network and automatically calibrating a pixel ratio.
Background
Most of domestic vehicle-mounted contact network monitoring device manufacturers use a fixed calibration tool to calibrate a final pixel ratio by means of a heating resistor or other special marks before calculating the geometric parameters of the contact network; when the device is installed, a specially-made calibration tool is fixed at the position of a pantograph of the car roof, the position of the mark is adjusted or the pixel ratio is calculated in a multi-point acquisition mode.
The calibration tool needs to be fixed at the pantograph of the car roof for measurement, operation needs to be carried out on the car roof during installation, potential safety hazards are large in the calibration process, periodic repeated calibration needs to be carried out in the system operation process to guarantee the measurement accuracy of geometric parameters, and meanwhile, the maintenance cost is increased for a vehicle operation unit.
Disclosure of Invention
Therefore, the invention provides a method for automatically calibrating a pixel ratio by calculating geometrical parameters of a contact network, and aims to solve the problems of low precision and high cost of the existing geometric parameter calibration of the contact network.
In order to achieve the above purpose, the invention provides the following technical scheme:
the invention discloses a method for calculating geometric parameters of a contact network and automatically calibrating a pixel ratio, which comprises the following steps:
identifying the characteristic points, selecting a plurality of fixed characteristic points, identifying the characteristic points through an identification device, and calculating to obtain a pixel ratio according to structural parameters given by a pantograph manufacturer or the actual distance of the manually and accurately measured characteristic points;
deep learning training, namely acquiring identification characteristic points aiming at a complex background and different pantograph models in the running process of a vehicle, automatically identifying and training the characteristic points, and constructing an identification model;
and automatic parameter correction, namely automatically identifying the characteristic points according to a set period, simultaneously calculating a pixel ratio, performing mean square error correction on the calculated pixel ratio and the previous pixel ratio, and writing the mean square error correction into a system parameter table.
Furthermore, the characteristic points are identified through an identification device, and sample training is performed in advance before the identification device leaves a factory, so that the identification capability of the basic characteristic points is achieved.
Further, the identification device is installed before the identification bit, measures the actual distance value of the feature point in advance, sets the actual distance value as L0, and writes the actual distance value into the system configuration parameter table.
Furthermore, the identification device detects the characteristic point first in the initial state of power-on start, calculates the initial value of the pixel ratio as P0, and writes the initial value into the system parameter table.
Further, the initial pixel ratio value is a ratio between the value identified by the initial feature point and the actual measurement result, the feature points are point1 and point2, the pixel coordinates are (x1, y1) and (x2, y2), and the distance between the two actually measured feature points is L0;
Figure BDA0003069256130000021
the subsequent pixel ratio P1-Pn is calculated by the same method as the initial pixel ratio P0.
Further, after the pixel ratio initial value P0 is calculated, geometric parameters are calculated through the pixel ratio initial value, and the geometric parameters are used for identifying the state of the pantograph, giving an alarm in real time and analyzing defects.
Furthermore, the recognition model improves the recognition degree through automatic training in the actual sample training and sampling processes, and the actual recognition result is continuously substituted into the recognition model to continuously optimize and update.
Further, the automatic parameter correction adopts an iterative updating mode, the mean square error of the pixel ratio parameter of the previous 30 days is taken every time to calculate the error range, and the calculation formula is as follows:
Figure BDA0003069256130000022
in the formula, q is a first error value, and P' is a pixel ratio of 30 days in history.
Further, after the first error value q is calculated, a mean square error calculation is performed on the pixel ratio parameter obtained on the current day to obtain a current measurement point error value Qn, and the calculation formula is as follows:
Figure BDA0003069256130000031
in the formula, Qn is the error value of the current measurement point, Pn is the pixel ratio of the current measurement point, when
Figure BDA0003069256130000032
When the Pn parameter is valid, the Pn parameter is replaced by P0 in a mean value mode, and the calculation process is as follows:
Figure BDA0003069256130000033
the invention has the following advantages:
the invention discloses a method for calculating geometric parameters of a contact network and automatically calibrating a pixel ratio, which simplifies a calibration process of a geometric parameter acquisition module of the contact network, adopts automatic calibration without manual intervention, greatly saves the product installation and debugging period, avoids safety accidents possibly occurring in the calibration process, and saves the maintenance cost because a product user does not need to calibrate periodically.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below. It should be apparent that the drawings in the following description are merely exemplary, and that other embodiments can be derived from the drawings provided by those of ordinary skill in the art without inventive effort.
The structures, ratios, sizes, and the like shown in the present specification are only used for matching with the contents disclosed in the specification, so as to be understood and read by those skilled in the art, and are not used to limit the conditions that the present invention can be implemented, so that the present invention has no technical significance, and any structural modifications, changes in the ratio relationship, or adjustments of the sizes, without affecting the effects and the achievable by the present invention, should still fall within the range that the technical contents disclosed in the present invention can cover.
Fig. 1 is a flowchart of a method for calculating geometric parameters of a catenary and automatically calibrating a pixel ratio according to an embodiment of the present invention;
Detailed Description
The present invention is described in terms of particular embodiments, other advantages and features of the invention will become apparent to those skilled in the art from the following disclosure, and it is to be understood that the described embodiments are merely exemplary of the invention and that it is not intended to limit the invention to the particular embodiments disclosed. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Examples
The embodiment discloses a method for calculating geometric parameters of a contact network and automatically calibrating a pixel ratio, which comprises the following steps:
identifying the characteristic points, selecting a plurality of fixed characteristic points, identifying the characteristic points through an identification device, and calculating to obtain a pixel ratio according to structural parameters given by a pantograph manufacturer or the actual distance of the manually and accurately measured characteristic points;
deep learning training, namely acquiring identification characteristic points aiming at a complex background and different pantograph models in the running process of a vehicle, automatically identifying and training the characteristic points, and constructing an identification model;
and automatic parameter correction, namely automatically identifying the characteristic points according to a set period, simultaneously calculating a pixel ratio, performing mean square error correction on the calculated pixel ratio and the previous pixel ratio, and writing the mean square error correction into a system parameter table.
The characteristic point identification is to perform directional identification on fixed characteristic points of the pantograph, and because the form of the pantograph and the head of the pantograph are basically kept fixed in the process of train moving, a certain number of fixed characteristic points can be selected for identification, which are generally larger than two characteristic points, and then the pixel ratio is calculated according to structural parameters given by a pantograph manufacturer or the actual distance of the manually accurately measured characteristic points.
The deep learning is a method for recognizing feature points under complex backgrounds and different pantograph models in the vehicle running process, the recognition degree is improved mainly through automatic training in the actual sample training and sampling processes, and the recognition rate is higher along with the improvement of the sample training amount, so that the recognition rate of the deep learning algorithm adopted by the product can reach more than 95%. The automatic correction parameters are that the pixel ratio parameters obtained by calculation can be automatically corrected in a certain period through continuous data acquisition and measurement in the running process of the product, and the accuracy of the calculated parameters is higher along with the improvement of the recognition rate.
The characteristic points are identified through an identification device, and sample training is performed in advance before the identification device leaves a factory to achieve the identification capability of the basic characteristic points; the identification device is installed before the identification position, measures the actual distance value of the characteristic point in advance, sets the actual distance value as L0, writes the actual distance value into the system configuration parameter table, detects the characteristic point firstly under the initial state of power-on starting, calculates the initial value of the pixel ratio as P0, and writes the initial value into the system parameter table.
Setting the initial pixel ratio value as the ratio between the initial characteristic point identification value and the actual measurement result, setting the characteristic points as point1 and point2, setting the pixel coordinates as (x1, y1) and (x2, y2), normally taking two characteristic points, if the number of the characteristic points is more than 2, respectively calculating and averaging every 2 points, and actually measuring the distance between the two characteristic points as L0;
Figure BDA0003069256130000051
the subsequent pixel ratio P1-Pn is calculated by the same method as the initial pixel ratio P0.
And after the pixel ratio initial value P0 is calculated, calculating geometric parameters through the pixel ratio initial value, wherein the geometric parameters are used for identifying the bow net state, alarming in real time and analyzing defects.
The recognition model improves the recognition degree through automatic training in the actual sample training and sampling processes, and the actual recognition result is continuously substituted into the recognition model to continuously optimize and update. In the actual detection process, different deviations can appear, the deviations need to be corrected, the corrected parameters are substituted into the recognition model, multiple rounds of training are carried out, and the recognition accuracy is improved.
The automatic parameter correction adopts an iterative updating mode, the mean square error of the pixel ratio parameter of the previous 30 days is taken every time to calculate the error range, and the calculation formula is as follows:
Figure BDA0003069256130000052
in the formula, q is a first error value, and P' is a pixel ratio of 30 days in history.
After the first error value q is calculated, performing mean square error calculation on the pixel ratio parameter obtained on the current day to obtain a current measurement point error value Qn, wherein the calculation formula is as follows:
Figure BDA0003069256130000061
in the formula, Qn is the error value of the current measurement point, Pn is the pixel ratio of the current measurement point, when
Figure BDA0003069256130000062
When the Pn parameter is valid, the Pn parameter is replaced by P0 in a mean value mode, and the calculation process is as follows:
Figure BDA0003069256130000063
the embodiment discloses a method for automatically calibrating a pixel ratio by calculating geometrical parameters of a contact network, which simplifies a calibration process of a geometrical parameter acquisition module of the contact network, adopts automatic calibration without manual intervention, greatly saves the product installation and debugging period, avoids safety accidents possibly occurring in the calibration process, and saves the maintenance cost because a product user does not need to calibrate periodically.
Although the invention has been described in detail above with reference to a general description and specific examples, it will be apparent to one skilled in the art that modifications or improvements may be made thereto based on the invention. Accordingly, such modifications and improvements are intended to be within the scope of the invention as claimed.

Claims (9)

1. A method for automatically calibrating a pixel ratio by calculating geometrical parameters of a contact network is characterized by comprising the following steps:
identifying the characteristic points, selecting a plurality of fixed characteristic points, identifying the characteristic points through an identification device, and calculating to obtain a pixel ratio according to structural parameters given by a pantograph manufacturer or the actual distance of the manually and accurately measured characteristic points;
deep learning training, namely acquiring identification characteristic points aiming at a complex background and different pantograph models in the running process of a vehicle, automatically identifying and training the characteristic points, and constructing an identification model;
and automatic parameter correction, namely automatically identifying the characteristic points according to a set period, simultaneously calculating a pixel ratio, performing mean square error correction on the calculated pixel ratio and the previous pixel ratio, and writing the mean square error correction into a system parameter table.
2. The method for calculating the automatic calibration pixel ratio of the geometric parameters of the overhead line system according to claim 1, wherein the feature points are identified by an identification device, and sample training is performed in advance before the identification device leaves a factory to achieve the identification capability of the basic feature points.
3. The method for calculating the automatic calibration pixel ratio of the geometric parameters of the overhead line system as claimed in claim 1, wherein the identification device is installed before the identification position to measure the actual distance value of the feature point in advance, set as L0, and write into the system configuration parameter table.
4. The method for automatically calibrating the pixel ratio in the geometric parameter calculation of the overhead line system of claim 1, wherein the identification device detects the characteristic point in the initial state of power-on start, calculates the initial value of the pixel ratio as P0, and writes the initial value into a system parameter table.
5. The method for automatically calibrating the pixel ratio in the geometric parameter calculation of the overhead line system as claimed in claim 4, wherein the initial value of the pixel ratio is the ratio between the value identified by the initial feature point and the actual measurement result, the feature points are point1 and point2, the pixel coordinates are (x1, y1) and (x2, y2), and the distance between the two actually measured feature points is L0;
Figure FDA0003069256120000011
the subsequent pixel ratio P1-Pn is calculated by the same method as the initial pixel ratio P0.
6. The method for automatically calibrating the pixel ratio in catenary geometric parameter calculation according to claim 4, wherein after the initial pixel ratio value P0 is calculated, the geometric parameters are calculated according to the initial pixel ratio value, and the geometric parameters are used for pantograph state identification, real-time alarm and defect analysis.
7. The method for calculating the automatic calibration pixel ratio of the geometric parameters of the overhead line system according to claim 1, wherein the recognition model improves the recognition degree through automatic training in the actual sample training and sampling processes, and the actual recognition result is continuously substituted into the recognition model to continuously optimize and update.
8. The method for calculating the automatic calibration pixel ratio of the geometric parameters of the overhead line system according to claim 1, wherein the automatic parameter correction adopts an iterative updating mode, the pixel ratio parameter of 30 days before each time is taken to perform mean square error calculation to calculate the error range, and the calculation formula is as follows:
Figure FDA0003069256120000021
in the formula, q is a first error value, and P' is a pixel ratio of 30 days in history.
9. The method for calculating the automatic calibration pixel ratio of the geometric parameters of the overhead line system according to claim 8, wherein after the first error value q is calculated, the mean square error of the pixel ratio parameters obtained on the same day is calculated to obtain the error value Qn of the current measurement point, and the calculation formula is as follows:
Figure FDA0003069256120000022
in the formula, Qn is the error value of the current measurement point, Pn is the pixel ratio of the current measurement point, when
Figure FDA0003069256120000023
When the Pn parameter is valid, the Pn parameter is replaced by P0 in a mean value mode, and the calculation process is as follows:
Figure FDA0003069256120000024
CN202110534974.8A 2021-05-17 2021-05-17 Method for automatically calibrating pixel ratio by calculating geometric parameters of contact network Active CN113405455B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202110534974.8A CN113405455B (en) 2021-05-17 2021-05-17 Method for automatically calibrating pixel ratio by calculating geometric parameters of contact network

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202110534974.8A CN113405455B (en) 2021-05-17 2021-05-17 Method for automatically calibrating pixel ratio by calculating geometric parameters of contact network

Publications (2)

Publication Number Publication Date
CN113405455A true CN113405455A (en) 2021-09-17
CN113405455B CN113405455B (en) 2023-03-07

Family

ID=77678721

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110534974.8A Active CN113405455B (en) 2021-05-17 2021-05-17 Method for automatically calibrating pixel ratio by calculating geometric parameters of contact network

Country Status (1)

Country Link
CN (1) CN113405455B (en)

Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103292701A (en) * 2013-06-24 2013-09-11 哈尔滨工业大学 Machine-vision-based online dimensional measurement method of precise instrument
CN103488976A (en) * 2013-09-17 2014-01-01 北京联合大学 Stop mark real-time detection and distance measurement method based on intelligent driving
CN103759658A (en) * 2014-01-27 2014-04-30 成都国铁电气设备有限公司 Method for detecting contact net geometrical parameters based on infrared image processing
CN103837087A (en) * 2013-06-04 2014-06-04 中国科学院遥感与数字地球研究所 Automatic pantograph detection method based on active shape model
CN105674880A (en) * 2016-01-25 2016-06-15 成都国铁电气设备有限公司 Geometric parameter measuring method and system for overhead lines based on binocular principle
WO2017113535A1 (en) * 2015-12-30 2017-07-06 完美幻境(北京)科技有限公司 Method and apparatus for geometric calibration of camera
CN107289858A (en) * 2017-07-06 2017-10-24 广州市九州旗建筑科技有限公司 The measurement apparatus and method of virtual ruler built in a kind of digital picture
JP2019133702A (en) * 2019-04-02 2019-08-08 日本電信電話株式会社 Template matching device, method and program

Patent Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103837087A (en) * 2013-06-04 2014-06-04 中国科学院遥感与数字地球研究所 Automatic pantograph detection method based on active shape model
CN103292701A (en) * 2013-06-24 2013-09-11 哈尔滨工业大学 Machine-vision-based online dimensional measurement method of precise instrument
CN103488976A (en) * 2013-09-17 2014-01-01 北京联合大学 Stop mark real-time detection and distance measurement method based on intelligent driving
CN103759658A (en) * 2014-01-27 2014-04-30 成都国铁电气设备有限公司 Method for detecting contact net geometrical parameters based on infrared image processing
WO2017113535A1 (en) * 2015-12-30 2017-07-06 完美幻境(北京)科技有限公司 Method and apparatus for geometric calibration of camera
CN105674880A (en) * 2016-01-25 2016-06-15 成都国铁电气设备有限公司 Geometric parameter measuring method and system for overhead lines based on binocular principle
CN107289858A (en) * 2017-07-06 2017-10-24 广州市九州旗建筑科技有限公司 The measurement apparatus and method of virtual ruler built in a kind of digital picture
JP2019133702A (en) * 2019-04-02 2019-08-08 日本電信電話株式会社 Template matching device, method and program

Also Published As

Publication number Publication date
CN113405455B (en) 2023-03-07

Similar Documents

Publication Publication Date Title
CN108161936B (en) Optimized robot calibration method and device
CN116227752B (en) Park facility management system based on Internet of things
CN108151766B (en) Positioning method of magnetic nails, positioning navigation error correction method of magnetic nails and positioning device
CN109978938A (en) A kind of pillow spring detection method based on machine vision
CN113405455B (en) Method for automatically calibrating pixel ratio by calculating geometric parameters of contact network
CN105759258A (en) Method and device for detecting distance through use of distance sensor
CN113256723B (en) Automatic detection method for pantograph lifting time and pantograph head displacement curve
CN111390911A (en) Manipulator position calibration system and calibration method
CN112528566A (en) Real-time air quality data calibration method and system based on AdaBoost training model
CN110455240B (en) Automatic vehicle paint film detection method and system
CN108226092B (en) Near infrared spectrum similarity-based model out-of-bound sample identification method
CN111695735A (en) Railway bow net real-time early warning method, system and device based on flow calculation
CN106767620B (en) A kind of sensor installation detection method of high-precision displacement measurement system
CN113251962B (en) Ultrasonic parking space compensation system based on machine learning
CN115238840A (en) Patrol robot positioning method and system, storage medium and intelligent terminal
CN115019537A (en) Intelligent navigation system applied to low-quality network condition
CN108871725B (en) Correction method for wind tunnel experiment reference static pressure
CN113869112A (en) Instrument automatic reading method and device based on machine vision
CN110021027B (en) Edge cutting point calculation method based on binocular vision
CN110515309B (en) Identification method for tracing source by dynamic simulation verification error of WAMS data power system
CN113587940A (en) Lane line checking method and system based on vehicle turning radius and vehicle
CN105865348A (en) Displacement measurement correction device and method
CN101419280A (en) Radar data fan number accounting and monitoring method
CN111457865A (en) Method and system for measuring straightness of band saw blade or steel band
CN111998792A (en) On-line detection method and on-line detection device for tire blank out-of-roundness of tire building machine

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