CN114295858B - Train speed acquisition and camera triggering method and device based on multi-source data fusion - Google Patents
Train speed acquisition and camera triggering method and device based on multi-source data fusion Download PDFInfo
- Publication number
- CN114295858B CN114295858B CN202111672028.6A CN202111672028A CN114295858B CN 114295858 B CN114295858 B CN 114295858B CN 202111672028 A CN202111672028 A CN 202111672028A CN 114295858 B CN114295858 B CN 114295858B
- Authority
- CN
- China
- Prior art keywords
- train
- speed
- measurement
- image
- magnetic steel
- 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
Landscapes
- Electric Propulsion And Braking For Vehicles (AREA)
- Length Measuring Devices By Optical Means (AREA)
Abstract
The invention discloses a train speed acquisition and camera triggering method and device based on multi-source data fusion, wherein the method comprises the following steps: the magnetic steel array and the image speed measuring module are adopted to collect train motion information, the train motion information is transmitted to the PLC for processing, the trigger frequency of the binocular camera is calculated in real time, the binocular camera is driven to collect train images in uniform scale, and train dynamic detection is achieved. According to the invention, the wheels of the train are accurately detected in real time through the binocular camera, and a wheel speed stereoscopic vision measurement model is established, so that the movement speed of the wheels of the train can be accurately measured; designing a train speed measurement mode based on a plurality of groups of magnetic steels, accurately measuring the train movement acceleration, and obtaining real-time speed information of train passing through by an integration mode; filtering by a Kalman filtering mode and carrying out weighted fusion to obtain the accurate speed of train movement; the invention improves the accuracy and reliability of train speed measurement, and enables the camera to meet the practical requirements of high-precision triggering and image acquisition in complex severe environments.
Description
Technical Field
The invention belongs to the technical field of rail transit safety detection, and particularly relates to a train speed acquisition and camera triggering method and device based on multi-source data fusion.
Background
On-line dynamic visual detection of key parts such as train bodies and wheels on the railway site becomes an important guarantee for ensuring safe operation of trains. However, factors such as complex and changeable field environment, train speed regulation, start-stop, guiding and the like always influence accurate detection of key parts of the train, and the main appearance is that the train speed cannot be accurately measured and the camera cannot accurately acquire related images. The existing magnetic steel-based speed measurement mode cannot meet the accurate detection of the train speed under the conditions of non-uniform passing of ultra-low speed, speed regulation, speed change and the like, and the magnetic steel-based speed measurement mode is easily interfered by the electromagnetic environment of the railway site, so that the speed measurement precision is low; the vision measurement method is widely applied to the fields of size measurement, attitude measurement and the like due to the characteristics of large range, high precision, non-contact and the like.
Disclosure of Invention
In order to solve the technical problems in the prior art, the invention aims to provide a train speed acquisition and camera triggering method and device based on multi-source data fusion.
In order to achieve the above purpose and achieve the above technical effects, the invention adopts the following technical scheme:
a train speed acquisition and camera triggering method based on multi-source data fusion comprises the following steps: based on the multi-sensor fusion measurement principle, the magnetic steel array and the image speed measurement module are adopted to collect train motion information, the train motion information is respectively transmitted to the PLC for analysis and processing, the trigger frequency of the binocular camera is calculated in real time, the binocular camera is driven to collect train images in uniform scale, and train dynamic detection is achieved.
Further, the train speed acquisition and camera triggering method based on multi-source data fusion comprises the following steps:
step one: establishing a three-dimensional vision measurement model of wheel speed
Based on a binocular stereoscopic vision measurement principle, detecting wheel areas in two groups of images acquired by a binocular camera through an image speed measurement module, matching, calculating three-dimensional information of characteristic points in a visual field of the binocular camera in real time, solving to obtain tangential running speed of a train, and constructing a wheel speed stereoscopic vision measurement model;
step two: establishing a vehicle speed measurement model based on a magnetic steel array
Arranging a plurality of groups of magnetic steels in the coming direction, wherein when a train passes through the magnetic steels, train wheels pass through the magnetic steels, calculating according to the passing time difference of the adjacent magnetic steels and the distance between the magnetic steels to obtain train running acceleration, obtaining real-time train running speed through integration, wherein the magnetic steel output measurement frequency is related to the train running speed, belongs to discrete sampling data, and is used for obtaining continuous train speed by utilizing real-time train running speed processing, constructing a train speed measurement model based on a magnetic steel array, and considering that the train stops when the triggering time interval of the adjacent magnetic steels is larger than a preset time threshold value;
step three: fusion speed measurement and triggering based on Kalman filtering mode
Filtering the measured speed by adopting a Kalman filtering mode, removing coarse errors to obtain a smooth speed curve, and obtaining a stable final speed value after weighted fusion:
after the magnetic steel detects an incoming train, an incoming train command is started, wheels sequentially pass through the magnetic steel, each magnetic steel detects train data and then transmits the train data to the PLC, an image speed measuring module detects the train data and then transmits the train data to the PLC, the PLC receives train information comprising a magnetic steel speed measuring value and an image speed measuring value, analyzes and processes the train information, obtains a final speed value after weighted fusion, generates high-frequency pulses, and triggers a binocular camera to acquire train images; and when the triggering quantity of all the magnetic steels is equal, judging that the vehicle passing is finished.
Further, the step of receiving train information including the magnetic steel speed measurement value and the image speed measurement value by the PLC and analyzing and processing the train information includes:
judging the speed effectiveness;
filtering the measurement data to reduce the influence of measurement noise;
setting a fixed sampling frequency of speed data according to the maximum speed and the allowable measurement error, and carrying out data interpolation and filling on each group of speed measurement data to unify the speed measurement data to the same sampling frequency;
and obtaining a final speed value according to the weighted fusion, generating high-frequency pulse, and triggering the binocular camera to acquire the current train image.
Further, in the train speed acquisition and camera triggering method based on multi-source data fusion, the step of judging the speed effectiveness comprises the following steps:
when the speed effectiveness judgment is carried out, when the passing speed V of the train is more than 0, calculating the displacement of the wheels of the train, then carrying out displacement judgment, and if the displacement change of the wheels of the train occurs, generating high-frequency pulses to trigger the binocular camera to acquire the current train image;
when the train passes through the speed v=0, no pulse is generated, and the binocular camera is not triggered;
when the passing speed V of the train is less than 0, calculating the displacement of the wheels of the train, judging the speed effectiveness again, and if the displacement change of the wheels of the train occurs, generating high-frequency pulses to trigger the binocular camera to acquire the current train image.
Further, the final velocity value v l The calculation formula of (2) is as follows:
v l ＝a 1 ·F 1 (v 1 )+a 2 ·F 2 (v 2 )
wherein a is 1 Representing the magnetic steel speed measurement weighting value F 1 (v 1 ) A represents a speed sequence of magnetic steel speed measurement, a 2 Representing the image speed measurement weighting value, F 2 (v 2 ) Representing a speed sequence of the image speed measuring module.
Further, the train passing speed V, the fixed sampling frequency n, the ideal pixel occupation number P and the target width D satisfy the following conditional expression:
the invention also discloses a device adopting the train speed acquisition and camera triggering method based on multi-source data fusion, which comprises on-track equipment and off-track equipment, wherein the on-track equipment and the off-track equipment are connected through a serial port and a network cable, the on-track equipment comprises an image speed measurement module and a magnetic steel array which are arranged on a train rail, the image speed measurement module comprises binocular cameras and an image speed calculation module, the surface surfaces of the binocular cameras are coaxially arranged on two sides of the rail, the binocular cameras are connected with the image speed calculation module, the magnetic steel array comprises a plurality of magnetic steels which are arranged at intervals, the off-track equipment comprises a PLC and a power supply module, the magnetic steel array and the image speed calculation module are respectively connected with the PLC, current train wheel images are acquired in real time through the binocular cameras based on a multi-sensor fusion measurement principle, the magnetic steel array and the image speed calculation module are adopted to complete acquisition of train motion information, the current train motion information is respectively transmitted to the PLC for analysis and processing, continuous and accurate motion information of the current train is output after filtering and weighting fusion, the triggering frequency of the binocular cameras is calculated in real time, and the acquisition of the train images is driven to complete even scale acquisition of the train images, and dynamic detection of the train is realized.
Compared with the prior art, the invention has the beneficial effects that:
1. based on a deep learning method, train wheels are accurately detected in real time through binocular cameras, a wheel speed stereoscopic vision measurement model is established, and the train wheel movement speed is accurately measured;
2. designing a train speed measurement mode based on a plurality of groups of magnetic steels, accurately measuring the train movement acceleration, and obtaining real-time speed information of train passing through by an integration mode;
3. the invention can improve the accuracy and reliability of train speed measurement, is particularly suitable for dynamic measurement of train speed during speed regulation, start-stop and guiding, and can enable a camera to meet the practical requirements of high-precision triggering and image acquisition in complex severe environments.
Drawings
FIG. 1 is a schematic diagram of the principle structure of the present invention;
FIG. 2 is a graph of wheel test alignment results according to the present invention;
FIG. 3 is a schematic diagram of the magnetic steel speed measurement of the present invention;
FIG. 4 is a graph of the analog tachometer data of the present invention before and after filtering;
FIG. 5 is a graph of the speed measurement and sampling period of the magnetic steel and binocular camera of the present invention;
FIG. 6 is a schematic diagram of a fusion estimation algorithm of the present invention;
fig. 7 is a flow chart of the speed calculation of the present invention.
Detailed Description
The present invention will be further described in detail with reference to the following examples, so that the advantages and features of the present invention can be more easily understood by those skilled in the art, and thus the scope of the present invention is more clearly and clearly defined.
The following presents a simplified summary of one or more aspects in order to provide a basic understanding of such aspects. This summary is not an extensive overview of all contemplated aspects, and is intended to neither identify key or critical elements of all aspects nor delineate the scope of any or all aspects. Its sole purpose is to present some concepts of one or more aspects in a simplified form as a prelude to the more detailed description that is presented later. It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the scope of the invention.
As shown in fig. 1-7, the train speed acquisition and camera triggering method based on multi-source data fusion is based on a multi-sensor fusion measurement principle, and the acquisition of train motion information is completed by adopting a magnetic steel array and an image speed measurement module, and the acquisition is respectively transmitted to a PLC for analysis and processing, so that the triggering frequency of a binocular camera is calculated in real time, and the binocular camera is driven to complete the acquisition of the uniform scale of train images, thereby realizing the dynamic detection of the train, and specifically comprising the following steps:
step one: establishing a three-dimensional vision measurement model of wheel speed
Based on the binocular stereoscopic vision measurement principle, detecting wheel areas in two groups of images acquired by the binocular camera 1 through an image speed measurement module, matching, calculating three-dimensional information of characteristic points in the field of view of the binocular camera 1 in real time, solving to obtain the tangential running speed of a train, and constructing a wheel speed stereoscopic vision measurement model;
step two: establishing a vehicle speed measurement model based on a magnetic steel array
Arranging a plurality of groups of magnetic steels 2 in the coming direction, wherein when a train passes through the magnetic steels 2, train running acceleration is obtained by calculating the passing time difference of adjacent magnetic steels 2 and the distance between the magnetic steels 2, real-time running speed of the train is obtained by integration, the output measurement frequency of the magnetic steels 2 is related to the train running speed, discrete sampling data are included, the real-time running speed of the train is utilized to process and obtain continuous train speed, a train speed measurement model based on a magnetic steel array is constructed, and when the triggering time interval of the adjacent magnetic steels 2 is larger than a preset time threshold, the train is considered to stop;
step three: fusion speed measurement and triggering based on Kalman filtering mode
Filtering the measured speed by adopting a Kalman filtering mode, removing coarse errors, and obtaining a stable final speed value v after weighted fusion l ：
v l ＝a 1 ·F 1 (v 1 )+a 2 ·F 2 (v 2 )
Wherein a is 1 Representing the magnetic steel speed measurement weighting value F 1 (v 1 ) A represents a speed sequence of magnetic steel speed measurement, a 2 Representing the image speed measurement weighting value, F 2 (v 2 ) The speed sequences representing the speed measurement of the image speed measurement module are obtained by testing the train speed; normalizing the data of the speed measured by the magnetic steel 2 to obtain a magnetic steel speed measurement weighted value, and normalizing the data measured by the image speed measurement module to obtain an image speed measurement weighted value;
after the magnetic steel 2 detects an incoming train, an incoming train command is started, wheels sequentially pass through the magnetic steel 2, each magnetic steel 2 detects train data and then transmits the train data to the PLC, an image speed measuring module detects the train data and then transmits the train data to the PLC, the PLC receives train information comprising a magnetic steel speed measuring value and an image speed measuring value, analyzes and processes the train information, obtains a final speed value after weighted fusion, generates high-frequency pulses, and triggers the binocular camera 1 to acquire train images; and when the triggering quantity of all the magnetic steels 2 is equal, judging that the vehicle passing is finished.
The steps of receiving train information including the magnetic steel speed measurement value and the image speed measurement value by the PLC and analyzing and processing the train information include:
first, a speed validity judgment is performed:
when the passing speed V of the train is more than 0, calculating the displacement of the wheels of the train, judging the displacement, and if the displacement of the wheels of the train changes, generating high-frequency pulses to trigger the binocular camera 1 to acquire the current train image;
when the train passes through the speed v=0, no pulse is generated, and the binocular camera 1 is not triggered;
when the passing speed V of the train is less than 0, calculating the displacement of the wheels of the train, judging the speed effectiveness again, and if the displacement change of the wheels of the train occurs, generating high-frequency pulses to trigger the binocular camera 1 to acquire the current train image;
secondly, filtering processing is carried out on each measurement data to reduce the influence of measurement noise;
finally, setting a speed data fixed sampling frequency according to the maximum passing speed and the allowable measurement error, carrying out data interpolation and filling on each group of speed measurement data, unifying the speed measurement data to the same sampling frequency, obtaining a final speed value according to weighted fusion, generating high-frequency pulse, and triggering the binocular camera 1 to acquire the current train image;
the passing speed V of the train, the fixed sampling frequency n of speed data, the ideal pixel number P and the target width D meet the condition:
the invention also discloses a train speed acquisition and camera triggering device based on multi-source data fusion, which comprises on-track equipment and off-track equipment, wherein the on-track equipment and the off-track equipment are connected through a serial port and a network cable, the on-track equipment comprises an image speed measurement module and a magnetic steel array which are arranged on a train rail, the image speed measurement module comprises a binocular camera 1 and an image speed calculation module 3, the surface faces of the binocular camera 1 and the image speed calculation module are coaxially arranged on two sides of the rail, the binocular camera 1 is connected with the image speed calculation module 3, the magnetic steel array comprises a plurality of magnetic steels 2 which are distributed at intervals, the off-track equipment comprises a PLC and a power supply module, the magnetic steel array and the image speed calculation module are respectively connected with the PLC, the current train wheel image is acquired in real time through the binocular camera 1 and transmitted to the image speed calculation module 3, the train speed is calculated, the acquisition of train motion information is completed through the magnetic steel array and the image speed calculation module 3, the image speed measurement module is respectively transmitted to the PLC for analysis and processing, the continuous and accurate current train motion information is output after the filtering and weighted fusion, the triggering frequency of the binocular camera 1 is calculated in real time, the binocular camera 1 is driven, the train image dimension is uniformly detected, and the train dynamic detection is realized.
Example 1
As shown in fig. 1-7, the invention discloses a train speed acquisition and camera triggering method based on multi-source data fusion, which is based on a multi-sensor fusion measurement principle, adopts a magnetic steel array, an image speed measurement module, a PLC (programmable logic controller) and the like to complete acquisition of train motion information, can realize accurate detection of train running speed and state under the conditions of train speed regulation and start-stop, calculates the triggering frequency of a high-speed binocular camera 1 in real time and drives the binocular camera 1 to complete acquisition of train images with uniform scale, and is suitable for train dynamic detection under the conditions of front line, warehouse-in and the like.
The train speed acquisition and camera triggering method based on multi-source data fusion disclosed by the embodiment specifically comprises the following steps:
step one: establishing a three-dimensional vision measurement model of wheel speed
The image speed measuring module is based on a binocular stereoscopic vision measuring principle, and the three-dimensional information of characteristic points in the visual field of the binocular camera 1 is calculated in real time by detecting and matching wheel areas in two groups of images acquired by the binocular camera 1, so that the tangential running speed of a train is obtained by solving, and a wheel speed stereoscopic vision measuring model is obtained; according to the invention, a near infrared light supplementing lamp is matched with the binocular camera 1, so that the aim of resisting outdoor parasitic light interference is fulfilled, and all-weather high-definition imaging of the vehicle body is ensured;
in this step, the target detection wheel region alignment based on the center net, in which the loss function is set as follows:
the offset error between the network output and the true value in the Heatmap is defined as:
wherein, alpha and beta are hyper-parameters of the focal loss, and N is the number of key points in the picture.
1.2 offset Loss
The offset Loss is designed, so that the trained network can effectively calculate the offset, accurately correct the position of the detection frame, and an offset Loss formula is set as follows:
where p is the true coordinates of the center point of the detection frame, p/R is the exact location area of the theoretical center point mapped to the feature map,is a network offset output profile, +.>Is the area into which the keypoints actually fall.
1.3, width-height Loss
Let the target wide-high area be defined as s k Consider the area into which the key point actually fallsIs defined asThen there are:
1.4, a total loss function,
The wheel target detection loss function is the sum of the three types of loss functions, and is defined as follows:
L det ＝L k +α size L size +α off L off
wherein alpha is size 、α off Respectively is L size Loss function weighting coefficient, L off A loss function weighting coefficient;
and carrying the train wheel set frame obtained through detection into a wheel speed stereoscopic vision measurement model, reconstructing to obtain the wheel movement displacement and speed, and further pushing out the forward movement speed of the train for fusion with the magnetic steel speed measurement data.
Step two: establishing a vehicle speed measurement model based on a magnetic steel array
At least six groups of magnetic steels 2 are arranged in the coming direction to form a magnetic steel array, so that wheels can pass the magnetic steels 2 when the train wheels pass, and according to the passing time difference t of the adjacent magnetic steels 2 and the distance L between the adjacent magnetic steels 2, the magnetic steel array is formed by the formula L=vt+1/2 at 2 Calculating to obtain train running acceleration a, and then obtaining real-time train running speed v through integration; when the triggering time interval of the adjacent magnetic steel 2 is larger than a preset time threshold value, the train is considered to stop; the output measurement frequency of the magnetic steel 2 is related to the train running speed, and belongs to discrete sampling data. Fig. 3 is a magnetic steel speed measurement principle, and the speed value between adjacent axles can be used for obtaining the speed of the continuous train through interpolation and fitting in the prior art, so that the speed measurement error is +/-2 cm.
Step three: fusion speed measurement and triggering based on Kalman filtering mode
Because the measuring speed is inevitably constrained by the motion condition of the train in the motion process, and the influence of outdoor parasitic light interference, dust, weather and appearance change of the train causes a great deal of noise to be overlapped, the measuring speed needs to be filtered, supplemented and smoothed to ensure the stability of the measuring speed. Kalman filtering (Kalman filtering) is an algorithm that uses a linear system state equation to optimally estimate the state of the system by inputting and outputting observed data through the system. The optimal estimate can also be seen as a filtering process, since the observed data includes the effects of noise and interference in the system. In the embodiment, a Kalman filtering mode is adopted to carry out filtering treatment on the measured speed, coarse errors are removed, and a smooth speed curve is obtained. The velocity of the magnetic steel 2 is recorded as v 1 The speed measured by the image speed measuring module is recorded as v 2 Filtered by a kalman filter algorithm, respectively. Fig. 4 is a graph of analog velocity measurement data before and after filtering, and it can be seen that coarse errors can be filtered out by using the kalman filtering method, and a smooth velocity curve can be obtained.
The magnetic steel 2 is used for measuring the speed most robustly, is little interfered by the external environment, but has low sampling frequency; image speed measuring module speed measuring frequency is influenced by object surface reliefThe sound is small; the magnetic steel 2 and the data measured by the image speed measuring module record unique time stamps, and after Kalman filtering is respectively carried out on each path of speed data, a stable final speed value v is obtained after weighted fusion l ：
v l ＝a 1 ·F 1 (v 1 )+a 2 ·F 2 (v 2 )
Wherein a is 1 Representing the magnetic steel speed measurement weighting value F 1 (v 1 ) A represents a speed sequence of magnetic steel speed measurement, a 2 Representing the image speed measurement weighting value, F 2 (v 2 ) The speed sequences representing the speed measurement of the image speed measurement module are obtained by testing the train speed;
starting the remote end, and starting an incoming command after the magnetic steel 2 detects an incoming; wheels sequentially pass through the magnetic steels 2, each magnetic steel 2 is detected to obtain train data and then is transmitted to the PLC, the image speed measuring module is used for detecting the train data and then is transmitted to the PLC, and the PLC receives train information including the magnetic steel speed measuring value and the image speed measuring value and analyzes and processes the train information:
firstly, judging the speed effectiveness;
secondly, filtering processing is carried out on each measurement data to reduce the influence of measurement noise;
furthermore, a fixed sampling frequency n of the speed data is set according to the maximum passing speed and the allowable measurement error, and a fixed sampling period F is determined: f= 2*n, set an acquisition target width D, an ideal pixel number P, a train passing speed V, a fixed sampling frequency n, and satisfy the following conditional expression:
in the embodiment, the passing speed v=30 km/h of the train, the acquisition target width d=10mm, the fixed sampling frequency n=833 Hz, and the fixed sampling period F approximately equals 2000Hz;
finally, carrying out data interpolation and filling according to each group of velocity measurement values, and unifying the velocity values of different measurements to the same sampling frequency; obtaining a final speed value according to weighted fusion, generating high-frequency pulses, and triggering the binocular camera 1 to acquire train wheel images; and when the triggering quantity of all the magnetic steels 2 is equal, judging that the vehicle passing is finished.
When the speed effectiveness judgment is carried out, when the passing speed V of the train is more than 0, calculating the displacement of the wheels of the train, then carrying out displacement judgment, and if the displacement change of the wheels of the train occurs, generating high-frequency pulses to trigger the binocular camera 1 to acquire the current train image;
when the train passes through the speed v=0, no pulse is generated, and the binocular camera 1 is not triggered;
when the passing speed V of the train is less than 0, the wheel displacement of the train is calculated, the speed effectiveness is judged again, if the wheel displacement of the train changes, a high-frequency pulse is generated, and the binocular camera 1 is triggered to collect the current train image.
The parts not specifically described are only needed to be obtained by adopting the prior art, and can be directly purchased in the market, and are not described in detail herein.
The foregoing description is only illustrative of the present invention and is not intended to limit the scope of the invention, and all equivalent structures or equivalent processes or direct or indirect application in other related arts are included in the scope of the present invention.
Claims (2)
1. The train speed acquisition and camera triggering method based on multi-source data fusion is characterized by comprising the following steps of:
step one: establishing a three-dimensional vision measurement model of wheel speed
Based on a binocular stereoscopic vision measurement principle, detecting wheel areas in two groups of images acquired by a binocular camera through an image speed measurement module, matching, calculating three-dimensional information of characteristic points in a visual field of the binocular camera in real time, solving to obtain tangential running speed of a train, and constructing a wheel speed stereoscopic vision measurement model;
step two: establishing a vehicle speed measurement model based on a magnetic steel array
Arranging a plurality of groups of magnetic steels in the coming direction, wherein when a train passes through the magnetic steels, train wheels pass through the magnetic steels, calculating according to the passing time difference of the adjacent magnetic steels and the distance between the magnetic steels to obtain train running acceleration, obtaining real-time train running speed through integration, wherein the magnetic steel output measurement frequency is related to the train running speed, belongs to discrete sampling data, and is used for obtaining continuous train speed by utilizing real-time train running speed processing, constructing a train speed measurement model based on a magnetic steel array, and considering that the train stops when the triggering time interval of the adjacent magnetic steels is larger than a preset time threshold value;
step three: fusion speed measurement and triggering based on Kalman filtering mode
And (3) carrying out filtering treatment on the measured speed by adopting a Kalman filtering mode, removing coarse errors, and obtaining a stable final speed value after weighted fusion:
after the magnetic steel detects an incoming train, an incoming train command is started, wheels sequentially pass through the magnetic steel, each magnetic steel detects train data and then transmits the train data to the PLC, an image speed measuring module detects the train data and then transmits the train data to the PLC, the PLC receives train information comprising a magnetic steel speed measuring value and an image speed measuring value, analyzes and processes the train information, obtains a final speed value after weighted fusion, generates high-frequency pulses, and triggers a binocular camera to acquire train images; when the triggering quantity of all the magnetic steels is equal, judging that the vehicle passing is finished;
the steps of receiving train information including the magnetic steel speed measurement value and the image speed measurement value by the PLC and analyzing and processing the train information include:
judging the speed effectiveness;
filtering the measurement data to reduce the influence of measurement noise;
setting a fixed sampling frequency of speed data according to the maximum speed and the allowable measurement error, and carrying out data interpolation and filling on each group of speed measurement data to unify the speed measurement data to the same sampling frequency;
obtaining a final speed value according to weighted fusion, generating high-frequency pulse, and triggering a binocular camera to acquire a current train image;
the step of judging the speed effectiveness comprises the following steps:
when the speed effectiveness judgment is carried out, when the passing speed V of the train is more than 0, calculating the displacement of the wheels of the train, then carrying out displacement judgment, and if the displacement change of the wheels of the train occurs, generating high-frequency pulses to trigger the binocular camera to acquire the current train image;
when the train passes through the speed v=0, no pulse is generated, and the binocular camera is not triggered;
when the passing speed V of the train is less than 0, calculating the displacement of the wheels of the train, judging the speed effectiveness again, and if the displacement change of the wheels of the train occurs, generating high-frequency pulses to trigger the binocular camera to acquire the current train image;
the final velocity value v l The calculation formula of (2) is as follows:
v l ＝a 1 ·F 1 (v 1 )+a 2 ·F 2 (v 2 )
wherein a is 1 Representing the magnetic steel speed measurement weighting value F 1 (v 1 ) A represents a speed sequence of magnetic steel speed measurement, a 2 Representing the image speed measurement weighting value, F 2 (v 2 ) Representing a speed sequence of the image speed measuring module for measuring the speed;
setting a fixed sampling frequency n of speed data according to the maximum passing speed and the allowable measurement error, and determining a fixed sampling period F: f= 2*n, set an acquisition target width D, an ideal pixel number P, a train passing speed V, a fixed sampling frequency n, and satisfy the following conditional expression:
2. the device adopting the train speed acquisition and camera triggering method based on multi-source data fusion as claimed in claim 1 is characterized by comprising on-track equipment and on-track equipment, wherein the on-track equipment and the on-track equipment are connected through serial ports and network wires, the on-track equipment comprises an image speed measurement module and a magnetic steel array which are arranged on a train rail, the image speed measurement module comprises binocular cameras and an image speed calculation module, the surfaces of the binocular cameras are coaxially arranged on two sides of the rail, the binocular cameras are connected with the image speed calculation module, the magnetic steel array comprises a plurality of magnetic steels which are distributed at intervals, the on-track equipment comprises a PLC and a power supply module, the magnetic steel array and the image speed calculation module are respectively connected with the PLC, current train wheel images are acquired through the binocular cameras based on a multi-sensor fusion measurement principle, the acquisition of train motion information is completed through the magnetic steel array and the image speed calculation module, the current train motion information is respectively transmitted to the PLC for analysis and processing, continuous and accurate motion information of the current train is output after filtering and weighting fusion, the triggering frequency of the binocular cameras is calculated in real time, and the acquisition of the images of the binocular cameras is completed, and the uniform scale of the train is achieved, and dynamic detection is achieved.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111672028.6A CN114295858B (en) | 2021-12-31 | 2021-12-31 | Train speed acquisition and camera triggering method and device based on multi-source data fusion |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111672028.6A CN114295858B (en) | 2021-12-31 | 2021-12-31 | Train speed acquisition and camera triggering method and device based on multi-source data fusion |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114295858A CN114295858A (en) | 2022-04-08 |
| CN114295858B true CN114295858B (en) | 2023-09-26 |
Family
ID=80975313
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111672028.6A Active CN114295858B (en) | 2021-12-31 | 2021-12-31 | Train speed acquisition and camera triggering method and device based on multi-source data fusion |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114295858B (en) |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN202057331U (en) * | 2011-01-14 | 2011-11-30 | 哈尔滨市科佳通用机电有限公司 | Train fault rail side image detection equipment based on multi-point speed measuring method |
| CN104573646A (en) * | 2014-12-29 | 2015-04-29 | 长安大学 | Detection method and system, based on laser radar and binocular camera, for pedestrian in front of vehicle |
| CN106627673A (en) * | 2016-10-27 | 2017-05-10 | 交控科技股份有限公司 | Multi-sensor fusion train positioning method and system |
| CN108981692A (en) * | 2018-06-14 | 2018-12-11 | 兰州晨阳启创信息科技有限公司 | It is a kind of based on inertial navigation/visual odometry train locating method and system |
| CN112556690A (en) * | 2020-11-09 | 2021-03-26 | 北京全路通信信号研究设计院集团有限公司 | Multi-source sensor fusion positioning method and device |
| CN112866673A (en) * | 2021-01-12 | 2021-05-28 | 北京格灵深瞳信息技术股份有限公司 | Image acquisition equipment, camera control method and device and train image acquisition system |
| CN113295161A (en) * | 2020-02-24 | 2021-08-24 | 中国航天科工飞航技术研究院（中国航天海鹰机电技术研究院） | Multisource information fusion's on-vehicle positioner that tests speed of hypervelocity maglev train |
| CN113838112A (en) * | 2021-09-24 | 2021-12-24 | 东莞市诺丽电子科技有限公司 | Trigger signal determining method and trigger signal determining system of image acquisition system |
-
2021
- 2021-12-31 CN CN202111672028.6A patent/CN114295858B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN202057331U (en) * | 2011-01-14 | 2011-11-30 | 哈尔滨市科佳通用机电有限公司 | Train fault rail side image detection equipment based on multi-point speed measuring method |
| CN104573646A (en) * | 2014-12-29 | 2015-04-29 | 长安大学 | Detection method and system, based on laser radar and binocular camera, for pedestrian in front of vehicle |
| CN106627673A (en) * | 2016-10-27 | 2017-05-10 | 交控科技股份有限公司 | Multi-sensor fusion train positioning method and system |
| CN108981692A (en) * | 2018-06-14 | 2018-12-11 | 兰州晨阳启创信息科技有限公司 | It is a kind of based on inertial navigation/visual odometry train locating method and system |
| CN113295161A (en) * | 2020-02-24 | 2021-08-24 | 中国航天科工飞航技术研究院（中国航天海鹰机电技术研究院） | Multisource information fusion's on-vehicle positioner that tests speed of hypervelocity maglev train |
| CN112556690A (en) * | 2020-11-09 | 2021-03-26 | 北京全路通信信号研究设计院集团有限公司 | Multi-source sensor fusion positioning method and device |
| CN112866673A (en) * | 2021-01-12 | 2021-05-28 | 北京格灵深瞳信息技术股份有限公司 | Image acquisition equipment, camera control method and device and train image acquisition system |
| CN113838112A (en) * | 2021-09-24 | 2021-12-24 | 东莞市诺丽电子科技有限公司 | Trigger signal determining method and trigger signal determining system of image acquisition system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114295858A (en) | 2022-04-08 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108974045A (en) | A kind of automatically walk track detector based on machine vision | |
| Zhan et al. | An accurate and efficient vision measurement approach for railway catenary geometry parameters | |
| CN108132025A (en) | A kind of vehicle three-dimensional outline scans construction method | |
| CN104021676A (en) | Vehicle positioning and speed measuring method based on dynamic video feature of vehicle | |
| CN109060821A (en) | Tunnel defect detection method and tunnel defect detection device based on laser detection | |
| CN106091951A (en) | A kind of municipal rail train wheel rim parameter on-line detecting system and method | |
| CN112119188B (en) | Method for controlling a set of one or more intervention tools mounted on a railway intervention vehicle | |
| CN109916305A (en) | Vehicle dimension measuring system and method based on optoelectronic induction and TOF technology | |
| CN114559988B (en) | Train speed, mileage, station and motion state measuring method, device and system | |
| CN109798850B (en) | Rail corrugation measuring method | |
| CN114719884A (en) | Attitude measurement precision evaluation method and application of inertial navigation system | |
| CN107340407A (en) | Train control system speed-position detection plan-validation method | |
| CN101264766A (en) | Orbit geometry parameter image measuring device and method for substituting chord wire method | |
| CN110926417B (en) | Vehicle-mounted railway tunnel detection system based on machine vision | |
| CN205991785U (en) | A kind of railway tunnel gauge dynamic detection system | |
| CN114295858B (en) | Train speed acquisition and camera triggering method and device based on multi-source data fusion | |
| CN103552570A (en) | Vehicle-mounted close-range photogrammetry method for detecting smoothness of railway track | |
| CN109017873A (en) | The method for carrying 2D laser radar scanning axle based on mobile detection robot | |
| CN102069821B (en) | Measuring method for non-contact steel rail sagging surface | |
| CN115540875B (en) | Method and system for high-precision detection and positioning of train vehicles in station track | |
| CN114219852A (en) | Multi-sensor calibration method and device for automatic driving vehicle | |
| CN106960440B (en) | Based on subject image characteristic point pixel space position combination control method | |
| CN111402171A (en) | Point cloud projection correction method based on tunnel general section | |
| CN109238149B (en) | Vehicle body attitude detection device and contact line dynamic offset detection system | |
| CN115451826B (en) | Photogrammetry method and device for geometric parameters of overhead contact system |
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 |