CN104239904A - Non-contact detection method for external outline of railway vehicle - Google Patents

Abstract

The invention discloses a non-contact detection method for the outer contour of a rail vehicle. The method comprises the steps of acquiring a digital photo of the outer contour of the vehicle under test by using a digital camera, analyzing, calculating and processing the digital photo, and drawing the outer contour of the vehicle under test. The graphics are compared with the standard outer contour graphics to realize the non-contact detection of the geometric dimensions of the outer contour of the vehicle under test; the method also includes the following steps: 1) equipment preparation, 2) establishment of a coordinate system, 3) setting of collection Points, 4) Acquiring digital photos, 5) Image processing and 6) Outputting results. The method of the present invention is scientific and reasonable, non-contact detection is performed on the outer contour of the rail vehicle, no damage is caused to the rail vehicle, no environmental pollution, strong anti-interference ability, high precision, real-time data collection, storage, analysis and processing can be completed, and the work efficiency is improved. high.

Description

Non-contact detection method for the outer contour of rail vehicles

technical field

The invention belongs to the technical field of railway vehicle safety detection, and in particular relates to a non-contact detection method for the outer contour of a rail vehicle based on machine vision detection technology.

Background technique

In recent years, with the rapid development of my country's high-speed railway and urban rail transit, as well as the continuous improvement of people's quality of life, coupled with the "going out" strategy being implemented by my country's rail transit equipment, higher requirements have been put forward for the manufacturing accuracy of the outline of new rail vehicles. Require. At present, according to the provisions of GB/T 16904 "Standard Gauge Railway Locomotive Vehicle Limit Inspection", for the measurement of the outer contour dimensions of rail vehicle bodies, my country still uses limit gauges. This method is time-consuming and laborious, and the detection accuracy is difficult to guarantee , therefore, the detection of the surface profile size of the new rail vehicle body has become a technical problem to be solved urgently. The limit gauge is divided into an upper gauge and a lower gauge. The upper gauge is composed of a steel structure frame, a central regulator, a leaf plate and a telescopic adjustment and positioning device, a two-way swing and a center reset mechanism, a non-contact sensor and a closed installation box, etc. Detect the size and geometric information of the top and side of the car body; the lower gauge is composed of three-dimensional adjustable support, leaf plate and support bearing, two-way swing and center reset device and other components. During the measurement, the inspected vehicle passes through the standard gauge railway rolling stock inspection section at a speed of less than 5km/h, and is measured by the expansion and contraction of the blade. Limit gauge detection mainly has the following shortcomings: The installation and adjustment process of the limit gauge is extremely complicated, and the operation requires manual participation, which is cumbersome and inefficient; The number of measurement points is limited, the reliability of the measurement results is poor, and the measurement data cannot be automatically saved, calculated and managed; Unable to obtain the external outline of the vehicle; When testing, it is necessary to reserve an inspection section of twice the length of the vehicle before and after the limit gauge, which increases the spatial range of the measurement.

Contents of the invention

The purpose of the present invention is to overcome the deficiencies of the prior art, aiming at the problems of low precision and low efficiency in the detection of the outer contour of newly built rail vehicles, based on machine vision theory and affine geometry principle, it provides a method that can realize the automatic measurement of the outer contour of rail vehicles. It is a non-contact detection method for the outer contour of rail vehicles that can analyze, evaluate and manage measurement data, reduce the number of test personnel, and improve the detection level.

The purpose of the present invention is to be achieved through the following technical solutions:

The method includes using a digital camera to obtain digital photos of the outer contour of the vehicle under test, analyzing, calculating and processing the digital photos, drawing the outer contour graphics of the tested vehicle, and comparing them with standard outer contour graphics to realize the detection of the measured vehicle. Non-contact detection of external contour geometry;

The method also includes the steps of:

1) Equipment preparation: set up a gantry, the vehicle under test can pass through the gantry, and the vehicle under test is equipped with a step-controlled traction device; several lighting lights and several detection devices are installed on the columns and top beams of the gantry; All lighting lights face the vehicle under test; each detection device includes a digital camera and a line light source, the line light source faces the vehicle under test vertically, and the digital camera faces the vehicle under test at a fixed angle; the irradiation area of the line light source falls into the digital camera within the imaging area; the centerlines of all line light sources are located in the same cross-section of the vehicle under test; the imaging areas of all digital cameras can be superimposed to completely cover the sides and top of the same cross-section of the vehicle under test; each detection device The control part of the traction device and the traction device are all electrically connected to the central control device. The central control device includes a microprocessor, a storage unit, an input interface and an output interface, and special software is installed, and the program controls the lighting, digital camera, line light source and traction device. , analyze, process and save data;

2) Establish a coordinate system: take the length direction of the measured vehicle as the X-axis, take the horizontal plane of the track where the tested vehicle is located as the Y-axis, and take the vertical centerline of the cross-section of the measured vehicle as the Z-axis. The O point of the XYZ coordinate system is located at the within the cross-section of the specified end of the vehicle under test;

3) Set the collection point: start from the O point of the XYZ coordinate system, for parts with small-sized parts on the outer surface of the vehicle under test, the shooting and collection distance along the X-axis direction is 3-10 mm, for other parts of the vehicle under test , the shooting and collection interval along the X-axis direction is 10-200 mm, and finally the M detection cross-sections are marked on the X-axis as M collection points, and their numbers and their coordinate values on the X-axis are saved;

4) Obtain digital photos: The number of collection points starts from 1 to the end of M. At a certain collection point, the lighting is always in working condition, and two digital photos of the background image and the target image are taken and saved successively: the online light source is turned off Get the background image under the condition of online light source to get the target image; get the digital photo by one of the following methods:

4A. The program controls the vehicle under test to travel along the X-axis intermittently, stop at each collection point and obtain digital photos:

4B. The program controls the vehicle under test to travel along the X-axis at a constant speed, and determines the time interval for obtaining photos according to the speed and the distance between two adjacent collection points on the X-axis. The program obtains digital photos according to the time interval:

4C. The program controls the vehicle under test to move continuously along the X axis. For parts with small-sized parts on the outer surface of the vehicle under test, the speed of the vehicle under test is 3-10 mm per second. For other parts of the vehicle under test, the speed of the vehicle under test is The speed of the vehicle is 10-200 mm per second. The time interval for obtaining photos is determined according to the speed and the distance between two adjacent collection points on the X-axis. The program obtains digital photos according to the time interval:

5) Image processing, using one of the following methods:

5A. Real-time processing: In the process of digital photo acquisition, the number of collection points starts from 1 to the end of M, and the corresponding background image and target image at each collection point are differentially processed in real time, and further processed by geometric calculation, in XYZ coordinates Draw the outer contour lines of the two sides and the top of the detection section in the system, and finally complete M outer contour lines;

5B. Centralized processing: After all the digital photos of the M collection points are taken, the numbers of the collection points start from 1 to the end of M, and the corresponding background images and target images at each collection point are differentially processed one by one, and further processed through geometric Calculation processing, drawing the outer contour lines of the two sides and the top of each detection section one by one in the XYZ coordinate system, a total of M outer contour lines;

6) Output results, depending on the comparison benchmarks, the following results can be output:

6A. Taking the standard limit as the comparison benchmark, compare the M outer contour lines one by one with the standard limit size, and output the coordinates and overrun size of the measured outer contour size of the vehicle that exceeds the limit;

6B. Taking the design drawing as a comparison benchmark, compare the M outer contour lines one by one with the size of the design drawing, and draw an error map of the measured outer contour size of the vehicle under test.

The detection device is also equipped with a digital camera, and the two digital cameras are symmetrically arranged on both sides of the center line of the line light of the line light source. The two digital cameras work synchronously. Two background images and two target images are obtained. In step "5) image processing", the two outer contour lines at the same collection point are averaged to improve detection accuracy.

The gantry can walk along the X axis under the program control of the central control device.

The digital camera is configured with a filter.

The detection device can be moved, adjusted and fixed along the Y axis and the Z axis.

Compared with the prior art, the present invention has the following advantages: the method is scientific and reasonable, non-contact detection is performed on the outer contour of the rail vehicle, no damage is caused to the rail vehicle, no environmental pollution, strong anti-interference ability, high precision, It can complete data collection, storage, analysis and processing in real time, with high work efficiency.

Description of drawings

Fig. 1 is the flowchart of detection method of the present invention;

Fig. 2 is a schematic diagram of the configuration of the equipment device according to an embodiment of the present invention;

Fig. 3 is a top view schematic diagram of Fig. 2, the top beam of the gantry is removed;

Fig. 4 is the A-A enlarged schematic diagram of Fig. 2;

In the figure: 1-track and foundation, 2-traveling device, 3-detection device, 4-gantry frame, 5-vehicle under test, 6-central control device, 7-digital camera, 8-line light source.

Detailed ways

The specific implementation of the detection method of the present invention will be further described in detail below in conjunction with the accompanying drawings.

Referring to the accompanying drawings: the method includes using a digital camera to obtain a digital photo of the outer contour of the tested vehicle 5, analyzing, calculating and processing the digital photo, drawing the outer contour figure of the tested vehicle 5, and comparing it with the standard outer contour figure , to realize the non-contact detection of the geometric dimensions of the external contour of the vehicle 5 under test;

The method also includes the steps of:

1) Equipment preparation: set up a gantry 4, the vehicle 5 to be tested can pass through the gantry 4, and the vehicle 5 to be tested is equipped with a step-controlled traction device; several lighting lights and some Each detection device 3; each lighting light is all facing the measured vehicle 5; each detection device 3 includes a digital camera 7 and a line light source 8, the line light source 8 faces the measured vehicle 5 vertically, and the digital camera 7 faces the measured vehicle 5 at a fixed angle The vehicle under test 5; the irradiated area of the line light source 8 falls in the imaging area of the digital camera 7; the centerlines of all the line light sources 8 are all located in the same cross section of the vehicle under test 5; the imaging areas of all the digital cameras 7 can be superimposed Completely cover both sides and the top of the same cross-section of the vehicle under test 5; the control parts of each detection device 3 and traction device are all electrically connected with the central control device 6, and the central control device 6 includes a microprocessor, a storage unit, and an input interface And output interface, install special software, program control lighting, digital camera 7, line light source 8 and traction device, collect, analyze, process and save data;

2) Establish a coordinate system: take the length direction of the tested vehicle 5 as the X axis, take the horizontal plane of the track where the tested vehicle 5 is located as the Y axis, and take the vertical centerline of the cross-section of the tested vehicle 5 as the Z axis, and the O of the XYZ coordinate system the point lies within the cross-section of the designated end of the vehicle under test 5;

3) Set collection points: starting from point O of the XYZ coordinate system, for parts with small-sized components on the outer surface of the measured vehicle 5, the shooting and collection distance along the X-axis direction is 3-10 mm, and for the parts of the measured vehicle 5 For other parts, the shooting and collection interval along the X-axis direction is 10-200 mm, and finally the M detection cross-sections are marked on the X-axis as M collection points, and their numbers and their coordinate values on the X-axis are saved;

4) Obtain digital photos: The number of collection points starts from 1 to the end of M. At a certain collection point, the lighting is always in working condition, and two digital photos of the background image and the target image are taken and saved successively: the online light source 8 is turned off Obtain the background image under the state, obtain the target image under the state that the online light source 8 is turned on; Obtain digital photo and select one of the following methods:

4A. The program controls the vehicle under test 5 to travel along the X-axis intermittently, stop at each collection point and obtain digital photos:

4B. The program controls the measured vehicle 5 to travel along the X-axis at a constant speed, and determines the time interval for obtaining photos according to the speed and the distance between two adjacent collection points on the X-axis. The program obtains digital photos according to the time interval:

4C. The program controls the tested vehicle 5 to move continuously along the X-axis. For parts with small-sized components on the outer surface of the tested vehicle 5, the tested vehicle 5 travels at a speed of 3-10 mm per second. For other parts of the tested vehicle 5 location, the traveling speed of the tested vehicle 5 is 10-200 mm per second, and the time interval for obtaining photos is determined according to the speed and the distance between two adjacent collection points on the X-axis, and the program obtains digital photos according to the time interval:

5) Image processing, using one of the following methods:

5A. Real-time processing: In the process of digital photo acquisition, the number of collection points starts from 1 to the end of M, and the corresponding background image and target image at each collection point are differentially processed in real time, and further processed by geometric calculation, in XYZ coordinates Draw the outer contour lines of the two sides and the top of the detection section in the system, and finally complete M outer contour lines;

5B. Centralized processing: After all the digital photos of the M collection points are taken, the numbers of the collection points start from 1 to the end of M, and the corresponding background images and target images at each collection point are differentially processed one by one, and further processed through geometric Calculation processing, drawing the outer contour lines of the two sides and the top of each detection section one by one in the XYZ coordinate system, a total of M outer contour lines;

6) Output results, depending on the comparison benchmarks, the following results can be output:

6A. Taking the standard limit as the comparison benchmark, compare the M outer contour lines one by one with the standard limit size, and output the coordinates and overrun size of the measured outer contour size of the measured vehicle 5 beyond the limit;

6B. Taking the design drawing as a comparison benchmark, compare the M outer contour lines one by one with the size of the design drawing, and draw an error map of the measured outer contour size of the vehicle 5 under test.

The detection device 3 is also equipped with a digital camera 7, and the two digital cameras 7 are symmetrically arranged on both sides of the center line of the line light of the line light source 8, and the two digital cameras 7 work synchronously. , two background images and two target images acquired at the same collection point, in step "5) image processing", the two outer contour lines at the same collection point are averaged to improve the detection accuracy.

The gantry 4 can walk along the X axis under the program control of the central control device 6 .

The digital camera 7 is configured with a filter.

The detection device 3 can be moved, adjusted and fixed along the Y axis and the Z axis.

Track and foundation 1: Two kinds of tracks are laid on the foundation, one is the track of the vehicle 5 to be tested, and the other is the track of the running device 2 of the gantry 4 .

Traveling device 2: installed at the bottom of the gantry 4, capable of driving the gantry 4 forward or backward.

Detection device 3 : an optical device for detecting the outer contour of the vehicle 5 under test, including a digital camera 7 and a line light source 8 .

Gantry 4: a carrier for installing lighting and detection devices 3.

Tested vehicle 5: a newly built vehicle, or a vehicle after overhaul.

Central control device 6: includes microprocessor, storage unit, input interface and output interface, installs special software, program controls lighting, digital camera 7, line light source 8 and traction device, collects, analyzes and processes and saves data.

Digital camera 7: A digital camera using existing technology with a resolution of not less than 768×576.

Line light source 8: adopt the line light source of the prior art.

Illumination light: adopt the illumination light of prior art.

Optical filter: adopt the optical filter of prior art.

In the following embodiment: the length of the tested vehicle 5 from a designated end to another end is 25000 mm.

Example 1:

1) Equipment preparation: set up a gantry 4, the vehicle 5 to be tested can pass through the gantry 4, and the vehicle 5 to be tested is equipped with a step-controlled traction device; several lighting lights and some Each detection device 3; each lighting light is all facing the measured vehicle 5; each detection device 3 includes a digital camera 7 and a line light source 8, the line light source 8 faces the measured vehicle 5 vertically, and the digital camera 7 faces the measured vehicle 5 at a fixed angle The vehicle under test 5; the irradiated area of the line light source 8 falls in the imaging area of the digital camera 7; the centerlines of all the line light sources 8 are all located in the same cross section of the vehicle under test 5; the imaging areas of all the digital cameras 7 can be superimposed Completely cover both sides and the top of the same cross-section of the vehicle under test 5; the control parts of each detection device 3 and traction device are all electrically connected with the central control device 6, and the central control device 6 includes a microprocessor, a storage unit, and an input interface And output interface, install special software, program control lighting, digital camera 7, line light source 8 and traction device, collect, analyze, process and save data;

2) Establish a coordinate system: take the length direction of the tested vehicle 5 as the X axis, take the horizontal plane of the track where the tested vehicle 5 is located as the Y axis, and take the vertical centerline of the cross-section of the tested vehicle 5 as the Z axis, and the O of the XYZ coordinate system the point lies within the cross-section of the designated end of the vehicle under test 5;

3) Set collection points: starting from point O of the XYZ coordinate system, for parts with small-sized parts on the outer surface of the vehicle 5 under test, the shooting and collection distance along the X-axis direction is 3 mm, and for other parts of the vehicle 5 under test , the shooting and collection interval along the X-axis direction is 100 mm, and finally the M detection cross-sections are marked on the X-axis as M collection points, and their numbers and their coordinate values on the X-axis are saved;

4) Obtain digital photos: The number of collection points starts from 1 to the end of M. At a certain collection point, the lighting is always in working condition, and two digital photos of the background image and the target image are taken and saved successively: the online light source 8 is turned off Obtain the background image under the state, obtain the target image under the state that the online light source 8 is turned on; Obtain the digital photo and select the following methods:

4A. The program controls the vehicle under test 5 to travel along the X-axis intermittently, stop at each collection point and obtain digital photos:

5) Image processing, real-time processing: In the process of digital photo acquisition, the number of collection points starts from 1 to the end of M, and the corresponding background image and target image at each collection point are differentially processed in real time, and further processed by geometric calculation. Draw the outer contour lines of the two sides and the top of the detection section in the XYZ coordinate system, and finally complete M outer contour lines;

6) To output the result, take the standard limit as the comparison benchmark, compare the M outer contour lines one by one with the standard limit size, and output the coordinates and overrun size of the measured outer contour size of the measured vehicle 5 beyond the limit;

Example 2:

It is basically the same as Embodiment 1, the difference is: the detection device 3 is also equipped with a digital camera 7, and the two digital cameras 7 are symmetrically arranged on both sides of the center line of the line light of the line light source 8, and the two digital cameras 7 work synchronously. In step "4) Acquiring digital photos", two background images and two target images are acquired at the same collection point. In step "5) Image processing", the two outer contour lines at the same collection point are averaged. In order to improve the detection accuracy.

Claims (9)

1. a rail vehicle exterior contour non-contact detection method, described method comprises the digital photograph adopting digital camera to obtain tested vehicle (5) exterior contour, logarithmic code photo carries out analytical calculation and process, draw out the outline figure of tested vehicle (5), compare with the outline figure of standard, realize carrying out contactless detection to tested vehicle (5) exterior contour physical dimension;

It is characterized in that: described method is further comprising the steps of:

1) equipment prepares: arrange a portal frame (4), can pass through tested vehicle (5) in portal frame (4), the draw-gear of tested vehicle (5) configuration step motion control; The column of portal frame (4) and top cross-bar install several houselights and several pick-up units (3); Each houselights is all towards tested vehicle (5); Each pick-up unit (3) comprises digital camera (7) and line source (8), line source (8) vertical plane to tested vehicle (5), digital camera (7) with fixed angle dip plane to tested vehicle (5); The irradiated region of line source (8) falls in the imaging area of digital camera (7); The center line of all line sources (8) is all positioned at the same xsect of tested vehicle (5); Superimposed both sides and the top that can cover the same cross-section of tested vehicle (5) completely, the imaging area of all digital cameras (7); The control section of each pick-up unit (3) and draw-gear is all electrically connected with central control unit (6), central control unit (6) comprises microprocessor, storage element, input interface and output interface, special software is installed, programmed control houselights, digital camera (7), line source (8) and draw-gear, gather, analyzing and processing preserve data;

2) coordinate-system is set up: with the length direction of tested vehicle (5) for X-axis, with tested vehicle (5) place track surface level for Y-axis, with the vertical center line of tested vehicle (5) xsect for Z axis, the O point of XYZ coordinate system is positioned at the xsect of the given end portion of tested vehicle (5);

3) collection point is set: from the O point of XYZ coordinate system, tested vehicle (5) outside surface is had to the position of small size parts, shooting, collecting spacing along X-direction is 3-10 millimeter, for other positions of tested vehicle (5), shooting, collecting spacing along X-direction is 10-200 millimeter, finally detect xsect using M to be labeled in X-axis as M collection point, and preserve their numberings and the coordinate figure in X-axis thereof;

4) obtain digital photograph: the numbering of collection point from 1 until M terminates, in certain collection point, houselights is in running order all the time, successively take and preserve background image and target image totally two digital photographs: background extraction image under the state that line source (8) is closed, obtains target image under the state that line source (8) is opened; Obtain digital photograph choosing one of with the following methods:

4A, programmed control tested vehicle (5) are advanced along X-axis with intermittent mode, and in each collection point, place stops and obtains digital photograph;

4B, programmed control tested vehicle (5) are advanced along X-axis at the uniform velocity mode, and according to speed and adjacent two collection points, the spacing in X-axis determines the time interval obtaining photo, and program obtains digital photograph according to the time interval;

4C, programmed control tested vehicle (5) are advanced along X-axis continuously, tested vehicle (5) outside surface is had to the position of small size parts, tested vehicle (5) gait of march is 3-10 millimeter per second, for other positions of tested vehicle (5), tested vehicle (5) gait of march is 10-200 millimeter per second, according to speed and adjacent two collection points, the spacing in X-axis determines the time interval obtaining photo, and program obtains digital photograph according to the time interval;

One of 5) image procossing, in the following ways:

5A, in real time process: in digital photograph acquisition process, the numbering of collection point from 1 until M terminates, the background image corresponding for each collection point place and target image carry out difference processing in real time, further across geometry computing, in XYZ coordinate system, draw out two sides of this detection sectional plane and the outer contour at top, finally complete M bar outer contour;

5B, to focus on: after the digital photograph of M collection point has all been taken, the numbering of collection point from 1 until M terminates, the background image corresponding for each collection point place and target image carry out difference processing one by one, further across geometry computing, two sides of each detection sectional plane and the outer contour at top is drawn out one by one, M bar outer contour altogether in XYZ coordinate system;

6) Output rusults, different according to the benchmark compared, following result can be exported:

6A, with standard gauge for benchmark, by M bar outer contour one by one compared with standard gauge size, export tested vehicle (5) actual measurement outer profile size exceed the coordinate at gauge position and the size that transfinites;

6B, be benchmark with design drawing, by M bar outer contour one by one compared with design drawing size, draw out the Error Graph of tested vehicle (5) actual measurement outer profile size.

2. method according to claim 1, it is characterized in that: described pick-up unit (3) also configures another digital camera (7), two digital cameras (7) are arranged in the center line symmetria bilateralis of the lines light of line source (8), two digital camera (7) synchronous workings, when step " 4) obtain digital photograph ", the background image that same collection point place obtains and each two of target image, when step " 5) image procossing ", two outer contours at same collection point place average process, to improve accuracy of detection.

3. method according to claim 1 and 2, is characterized in that: described portal frame (4) can be walked along X-axis under the programmed control of central control unit (6).

4. method according to claim 1 and 2, is characterized in that: described digital camera (7) configuration optical filter.

5. method according to claim 3, is characterized in that: described digital camera (7) configuration optical filter.

6. method according to claim 1 and 2, is characterized in that: described pick-up unit (3) can move adjustment along Y-axis and Z axis and fix.

7. method according to claim 3, is characterized in that: described pick-up unit (3) can move adjustment along Y-axis and Z axis and fix.

8. method according to claim 4, is characterized in that: described pick-up unit (3) can move adjustment along Y-axis and Z axis and fix.

9. method according to claim 5, is characterized in that: described pick-up unit (3) can move adjustment along Y-axis and Z axis and fix.