Track inspection system and inspection vehicle

Abstract

The utility model discloses a track inspection system and an inspection vehicle. The system comprises an image acquisition unit, a positioning synchronization unit, a switch and an image processing unit. The image acquisition unit, the switch and the image processing unit are sequentially connected, and the positioning synchronization unit is respectively connected with the image acquisition unit and the switch. And the positioning synchronization unit sends a trigger signal to the image acquisition unit to trigger the image acquisition unit to obtain 3D track data. And the positioning synchronization unit sends the mileage position data at the triggering moment to the image processing unit through the switch. 3D track data acquired by the image acquisition unit is transmitted to the image processing unit through the switch. According to the invention, the technical problem of high rail surface damage false detection rate due to the fact that the inspection system only adopts the rail two-dimensional gray level image to inspect the rail state can be solved.

Description

Technical Field

[0001] The present application relates to the technical field of railway engineering machinery, and in particular to a track inspection system and an inspection vehicle for guiding daily line maintenance. Background Art

[0002] Railway operation safety has always been a major issue related to economic development and personal safety. Tracks are the foundation of railway transportation. With the high-speed, heavy-load, and high-density operation of locomotives, locomotive wheels will wear the tracks when running on the rails. When the wear exceeds a certain limit, on the one hand, it will increase the contact area with the locomotive wheel tread, increase the running resistance, and have a great impact on the driving safety of the locomotive; on the other hand, the locomotive vehicle will sway violently, seriously affecting the comfort of passengers' travel, and more seriously, it will cause damage to fasteners and sleepers. If the disease is not treated for a long time, it will further expand and endanger driving safety. Therefore, it is necessary to regularly and timely detect the wear of track facilities and rail surfaces to ensure driving stability and safety.

[0003] At present, the railway engineering section's maintenance work along the railway mainly focuses on the maintenance of rails, switches, sleepers, fasteners and ballast, including complete and continuous recording of high-definition track status, identification of defects such as scratches on the rail surface, abnormal fasteners, and track plate cracks, to ensure the normal operation of the railway line. Previously, my country has always used manual inspections to check track status, and usually conducts them at night with a short window of time. This not only consumes a lot of manpower, material and financial resources, but also has a high rate of missed inspections and false inspections, and the inspection results are greatly affected by the subjective influence of the staff. In addition, because this method takes a long time to detect once, a line can often only be inspected once a month, which greatly increases the probability of accidents.

[0004] The main characteristics of railway engineering inspection work at present are:

[0005] 1) Manual inspection method. It requires a lot of people, has low efficiency and high labor intensity. Its inspection effect is directly related to the experience and sense of responsibility of the inspectors.

[0006] 2) Two-dimensional grayscale image detection method. The method is single and the false detection rate is high when using two-dimensional grayscale images for image recognition. It cannot accurately identify the oil and water stains on the track surface, the looseness of fastener bolts, or the ballast density inspection. The reliability is low.

[0007] In the prior art, the following documents are mainly related to this application:

[0008] Prior art 1 is a Chinese utility model patent applied by Chengdu Yunda Technology Co., Ltd. on August 26, 2019 and announced on April 24, 2020, with the announcement number CN210402469U. The utility model discloses a vehicle-mounted intelligent track inspection system, including off-board equipment, on-board equipment and vehicle depot equipment. The off-board equipment includes a detection beam arranged at the bottom of the vehicle, on which an electronic tag reader, a synchronous trigger device electrically connected to the electronic tag reader and at least three groups of high-definition imaging groups electrically connected to the synchronous trigger device are arranged; the on-board device includes a detection host, a front-end data processing device connected to the detection host for signal connection, and a wireless communication device and a wired communication device connected to the detection host for signal connection, and the electronic tag reader is connected to the detection host for signal connection; the information output by the distance encoder, the synchronous trigger device, etc. is received by the high-definition imaging group and modulated and processed, and then transmitted to the detection host for analysis and processing, which greatly improves the efficiency of track inspection and reduces the waste of manpower.

[0009] Prior art 2 is a Chinese invention application filed by Shanghai Xintie Electromechanical Technology Co., Ltd. on July 18, 2017, and published on November 21, 2017, with publication number CN107364466A. The invention discloses a non-contact track inspection system for rail vehicles, including a CCD high-speed line array camera device for photographing images of rails, fasteners, rail plates and roadbeds, and an image analysis and processing system for analyzing and processing images. The CCD high-speed line array camera device is installed on a suspension mounting seat mounted on a maintenance track inspection vehicle or a rail flaw detection vehicle. The invention solves the problems of single means, low reliability, and low work efficiency of existing detection methods, and the detection work is automated, informationized, continuous and precise, with small workload and high efficiency, reducing operating costs, and can provide high-quality comprehensive evaluation results. It can timely grasp the quality status of the track, correctly guide line maintenance and repair, ensure road transportation safety, and meet the needs of today's high-speed railway and urban rail transit development.

[0010] Prior art 3 is a Chinese invention application filed by Guangzhou Metro Group Co., Ltd. and Chengdu Mingju Technology Co., Ltd. on December 12, 2018, and published on April 19, 2019, with publication number CN109658397A. The invention discloses a track inspection method and system, which includes: obtaining image information and location information of train tracks; intercepting images of regions of interest from image information; inputting images of regions of interest into a deep learning network; and when the images of regions of interest are identified as abnormal, sending images of regions of interest and location information to a server and vehicle terminal monitoring device to implement track inspection. The invention uses intelligent identification of track defects to improve the recognition rate of track defects.

[0011] Prior art 4 is a Chinese utility model patent applied by the Institute of Science and Technology of China Railway Shanghai Bureau Group Co., Ltd. on January 30, 2013, and published on August 28, 2013, with the publication number CN203165044U. The utility model discloses a vehicle-mounted intelligent track inspection system, including a sensing module, a data acquisition and storage module, a data analysis module and a power module. The data acquisition and storage module is connected to and controls the sensing module to collect data, the data intelligent analysis module is connected to and receives the data in the data acquisition and storage module, and the power module is connected to the sensing module, the data acquisition and storage module, and the data analysis module respectively. The utility model can detect the falling off and skewing of track fasteners of high-speed railway ballastless tracks and existing line ballasted tracks; damage to the rail surface; abnormal distribution of light bands on the rail surface; mud slurry on the roadbed and the remains on the track plate; and cracks on the track plate. It has a simple structure, can detect more defect items under the same number of cameras, has high detection accuracy, is easy to install, use and maintain, and has a high cost performance.

[0012] The technical solutions of the above-mentioned prior art all adopt the two-dimensional imaging method, and do not collect three-dimensional information of the track. Since only the two-dimensional grayscale image of the track is used for image recognition, it is impossible to clearly identify oil stains and water stains on the track surface. Therefore, the existing system has a high false detection rate of rail surface damage, and the existing system cannot identify loose spikes and perform ballast density inspection through two-dimensional images. Utility Model Content

[0013] In view of this, the purpose of the present application is to provide a track inspection system and an inspection vehicle to solve the technical problem that the existing system only uses two-dimensional grayscale images of the track to inspect the track status, resulting in a high false detection rate of rail surface damage.

[0014] In order to achieve the above-mentioned utility model purpose, the present application specifically provides a technical implementation scheme of a track inspection system, including: an image acquisition unit, a positioning synchronization unit, a switch and an image processing unit. The image acquisition unit, the switch and the image processing unit are connected in sequence, and the positioning synchronization unit is connected to the image acquisition unit and the switch respectively. The positioning synchronization unit sends a trigger signal to the image acquisition unit to trigger the image acquisition unit to acquire 3D track data. The positioning synchronization unit sends the mileage position data at the triggering moment to the image processing unit through the switch. The 3D track data acquired by the image acquisition unit is transmitted to the image processing unit through the switch.

[0015] Furthermore, the system also includes a storage unit connected to the image processing unit, and the storage unit stores data output by the image processing unit.

[0016] Furthermore, the positioning synchronization unit includes a rotary encoder, a positioning synchronization mainboard and a tag reader. When the inspection vehicle is running, the rotary encoder outputs a pulse signal, the positioning synchronization mainboard receives the pulse signal from the encoder, and sends a trigger signal to the image acquisition unit at a set mileage interval to trigger the image acquisition unit to collect track data. The positioning synchronization mainboard sends the mileage position data at the time of triggering to the image processing unit through the switch. The tag reader is used to obtain the mileage information corresponding to a certain set position after the inspection vehicle passes the position.

[0017] Furthermore, the rotary encoder is installed on the wheel axle of the inspection vehicle.

[0018] Furthermore, the image acquisition unit is installed at the rear of the inspection vehicle.

[0019] Furthermore, the tag reader is arranged close to the image acquisition unit.

[0020] Furthermore, the positioning synchronization mainboard, switch, image processing unit and storage unit are all arranged in the middle of the body of the inspection vehicle.

[0021] Furthermore, the image processing unit includes a data acquisition and storage module, a data playback and historical data analysis module and an image recognition module.

[0022] Furthermore, the image acquisition unit is used to acquire the laser light strip image modulated by the height of the object surface, and then calculate the grayscale value and depth value of the object surface, and finally obtain the grayscale image and depth image of the object surface.

[0023] Furthermore, the image acquisition unit includes a left 3D inspection component and a right 3D inspection component for respectively acquiring left and right track data. The left 3D inspection component and the right 3D inspection component shoot vertically downward and acquire track infrastructure data including left and right rails, fasteners, sleepers and track plates.

[0024] Furthermore, the left 3D inspection component and the right 3D inspection component both include a line laser, a short-focus lens, a filter, and a 3D camera. The line laser emits a line laser of a specific wavelength to the track surface, and the line laser is highly modulated by the surface of the object to form a laser light strip, which is collected by the 3D camera after passing through the short-focus lens and the filter to form a laser light strip image, and then the grayscale value and depth value of the object surface are calculated, and finally the grayscale image and depth image of the object surface are obtained.

[0025] Furthermore, the system also includes a mileage tag set on the track line, and the tag reader is used to obtain the mileage information corresponding to a mileage tag after the inspection vehicle passes by the mileage tag.

[0026] Furthermore, the mileage labels are set at the starting positions of several mileage intervals divided by the track line.

[0027] The present application also specifically provides a technical implementation solution for a patrol inspection vehicle, the patrol inspection vehicle comprising: the track patrol inspection system as described above.

[0028] By implementing the technical solutions of the track inspection system and inspection vehicle provided by the present application, the following beneficial effects are achieved:

[0029] (1) The track inspection system and inspection vehicle of the present application establish three-dimensional track information and surface image information by acquiring grayscale images, depth images and mileage data of the track line, and use image recognition technology to accurately identify and locate abnormal positions of the track. This can solve the technical problem that the existing system only uses two-dimensional grayscale images of the track for track status inspection, and has a high false detection rate of rail surface damage;

[0030] (2) The rail inspection system and inspection vehicle of the present application realize the 3D rail inspection for qualitative identification and quantitative measurement of visual defects of rail infrastructure. It can not only qualitatively identify damage, but also quantitatively analyze damage, and can evaluate the development of damage in combination with historical information to guide workers to perform maintenance;

[0031] (3) The track inspection system and inspection vehicle of the present application are capable of performing 3D detailed modeling and comprehensive three-dimensional inspection of the track surface. The inspection objects include track facilities such as rails, fasteners, sleepers, track plates and induction plates, and can realize the measurement and intuitive three-dimensional display of track diseases, foreign body detachment, etc. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the prior art descriptions. Obviously, the drawings described below are only some embodiments of the present application, and for ordinary technicians in this field, other embodiments can be obtained based on these drawings without creative work.

[0033] Figure 1 It is a schematic diagram of the system principle of a specific embodiment of the track inspection system of the present application;

[0034] Figure 2 This is a system structure block diagram of a specific embodiment of the track inspection system of the present application;

[0035] Figure 3 This is a block diagram of the structure of a 3D inspection component of an image acquisition unit in a specific embodiment of the track inspection system of the present application;

[0036] Figure 4This is a block diagram of the structural composition and connection structure of the positioning synchronization unit in a specific embodiment of the track inspection system of the present application;

[0037] Figure 5 This is a block diagram of the structure of an image processing unit in a specific embodiment of the track inspection system of the present application;

[0038] Figure 6 This is a flowchart of a specific embodiment of the track inspection method based on the system of the present application;

[0039] Figure 7 It is a structural schematic diagram of a specific embodiment of the inspection vehicle of the present application;

[0040] In the figure: 1-image acquisition unit, 2-positioning synchronization unit, 3-switch, 4-image processing unit, 5-storage unit, 10-inspection vehicle, 11-left 3D inspection component, 12-right 3D inspection component, 101-line laser, 102-short focus lens, 103-filter, 104-3D camera, 20-vehicle body, 21-rotary encoder, 22-positioning synchronization mainboard, 23-label reader, 24-mileage label, 41-data acquisition and storage module, 42-data playback and historical data analysis module, 43-image recognition module. DETAILED DESCRIPTION

[0041] In order to make the purpose, technical scheme and advantages of the embodiments of the present application clearer, the technical scheme in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0042] As attached Figure 1 To Attachment Figure 7 As shown, a specific embodiment of the track inspection system and the inspection vehicle of the present application is given, and the present application is further explained below in conjunction with the drawings and specific embodiments.

[0043] Example 1

[0044] As attached Figure 1 and attached Figure 2As shown, an embodiment of the track inspection system of the present application specifically includes: an image acquisition unit 1, a positioning synchronization unit 2, a switch 3 and an image processing unit 4. The image acquisition unit 1, the switch 3 and the image processing unit 4 are connected in sequence, and the positioning synchronization unit 2 is connected to the image acquisition unit 1 and the switch 3 respectively. The positioning synchronization unit 2 sends a trigger signal to the image acquisition unit 1, triggering the image acquisition unit 1 to acquire 3D (i.e. three-dimensional) track data. The positioning synchronization unit 2 sends the mileage position data at the triggering moment to the image processing unit 4 through the switch 3. The 3D track data acquired by the image acquisition unit 1 is transmitted to the image processing unit 4 through the switch 3. The track inspection system also includes a storage unit 5 connected to the image processing unit 4, and the storage unit 5 stores the data output by the image processing unit 4.

[0045] In terms of system hardware composition, the image acquisition unit 1 is mainly composed of two left and right 3D inspection components that shoot vertically downward, wherein the 3D inspection components are mainly composed of a line laser 101, a short-focus lens 102, a filter 103 and a 3D camera. The image processing unit 4 further adopts a high-performance image processing host, with a high-performance GPU (Graphics Pprocessing Unit, the abbreviation of graphics processor). The storage unit 5 further adopts a disk array with a storage capacity of not less than 40TB. The positioning synchronization unit 2 is mainly composed of a rotary encoder 21, a positioning synchronization mainboard 22, a label reader 23 and a mileage label 24. The switch 3 further adopts a high-performance 10 Gigabit switch.

[0046] The track inspection system described in the specific embodiment of the present application mainly utilizes line structured light imaging technology, collects the laser light strip image modulated by the height of the object surface, calculates the depth value and grayscale value of each point on the object surface on the laser line, and obtains the mileage position of the laser line in combination with the trigger signal output when the detection platform (i.e., the inspection vehicle 10) moves, and then obtains the complete track cross-sectional profile information at the trigger position, and finally transmits it to the image processing unit 4 via Ethernet, obtains continuous track grayscale images and depth images and stores them in the disk array (i.e., the storage unit 5). This embodiment can realize the high-speed acquisition of two-dimensional images and three-dimensional data of track infrastructure such as rails, fasteners, sleepers, roadbeds, induction plates, etc., thereby realizing qualitative identification and quantitative measurement of visible track diseases.

[0047] The image acquisition unit 1 is used to acquire the laser light strip image modulated by the height of the object surface, and then calculate the grayscale value and depth value of the object surface, and finally obtain the grayscale image and depth image of the object surface. The image acquisition unit 1 further includes a left 3D inspection component 11 and a right 3D inspection component 12 for respectively acquiring the track data on the left and right sides. The left 3D inspection component 11 and the right 3D inspection component 12 shoot vertically downward and acquire the track infrastructure data including the left and right rails, fasteners, sleepers and track plates.

[0048] As attached Figure 3 As shown, the left 3D inspection component 11 and the right 3D inspection component 12 further include a line laser 101, a short-focus lens 102, a filter 103 and a 3D camera 104. The 3D camera 104 is used to collect three-dimensional information and grayscale information of the object surface. The line laser 101 emits a line laser of a specific wavelength to the track surface. The line laser is highly modulated by the surface of the object to form a laser light strip. The laser light strip is collected by the 3D camera 104 after passing through the short-focus lens 102 and the filter 103 to form a laser light strip image. The depth value and grayscale value of each point on the object surface on the laser line are calculated, and finally the grayscale image data and depth image data of the object surface are obtained. The image acquisition unit 1 is connected to the switch 3 through two Gigabit Ethernet interfaces and finally connected to the image processing unit 4, and the image data collected by the left and right 3D inspection components are transmitted to the image processing unit 4 for subsequent image processing and historical data analysis.

[0049] As attached Figure 4 As shown, the positioning synchronization unit 2 further includes a rotary encoder 21, a positioning synchronization mainboard 22 and a tag reader 23. When the inspection vehicle 10 is running, the rotary encoder 21 outputs a pulse signal, the positioning synchronization mainboard 22 receives the pulse signal from the encoder 21, and sends a trigger signal to the image acquisition unit 1 at a set mileage interval to trigger the image acquisition unit 1 to collect track data. The positioning synchronization mainboard 22 sends the mileage position data at the time of triggering to the image processing unit 4 through the switch 3. The tag reader 23 is used to obtain the mileage information corresponding to a certain set position after the inspection vehicle 10 passes the position.

[0050] The rotary encoder is installed on the wheel axle of the inspection vehicle 10. The image acquisition unit 1 is installed at the rear of the body 20 of the inspection vehicle 10. The tag reader 23 is arranged near the image acquisition unit 1. The track inspection system also includes a mileage tag 24 arranged on the track line, and the tag reader 23 is used to obtain the mileage information corresponding to the mileage tag 24 after the inspection vehicle 10 passes through the mileage tag 24. The mileage tag 24 is set at the starting position of several mileage intervals divided by the track line. The positioning synchronization mainboard 22, the switch 3, the image processing unit 4 and the storage unit 5 are all arranged in the middle of the body 20 of the inspection vehicle 10. As shown in the attached Figure 5As shown, the image processing unit 4 further includes a data acquisition and storage module 41 , a data playback and historical data analysis module 42 and an image recognition module 43 .

[0051] The image processing unit 4 is located in the middle of the body 20 of the inspection vehicle 10. The image processing unit 4 is a high-performance image processing host, which is responsible for processing image information and packaging and storing images, as well as unpacking the images of the hard disk and automatically identifying the images according to the different target areas of the corresponding camera, and obtaining the state and damage of the target object. The host of the image processing unit 4 has a gigabit network port, which can stably connect the image signal. The gigabit network interface has a fast transmission speed and a long transmission distance, and the maximum transmission distance can exceed 100m. The storage unit 5 is located in the middle of the body 20 of the inspection vehicle 10. The storage unit 5 is a disk array composed of several hard disks, with a total capacity of not less than 40TB, which is responsible for the long-term storage of the collected track data.

[0052] The image processing unit 4 is a 3D inspection host, which is responsible for running the detection software to realize data acquisition and storage, data playback and historical data analysis, and image recognition. The storage unit 5 is a disk array, which is responsible for storing the packaged track grayscale image data and depth image data. The positioning synchronization unit 2 is a positioning synchronization system, which is responsible for sending a trigger signal to the image acquisition unit 1, triggering the image acquisition unit 1 to collect track data, and at the same time recording the mileage position at the time of collection and sending it to the image processing unit 4. The switch 3 is mainly responsible for connecting each unit for data transmission. The details of each unit are introduced as follows.

[0053] The positioning synchronization unit 2 is a positioning synchronization system, which is located in the middle of the body 20 of the inspection vehicle 10. The positioning synchronization unit 2 is mainly composed of a rotary encoder 21, a positioning synchronization mainboard 22, a label reader 23 and a mileage label 24. Among them, the rotary encoder 21 is installed at the wheel of the inspection vehicle 10. When the inspection vehicle 10 is running, the rotary encoder 21 outputs pulse signals at equal angles (equal distances). The positioning synchronization mainboard 22 receives the pulse signal from the rotary encoder 21, calculates the mileage at the current position in combination with the starting mileage, and sends a trigger signal to the image acquisition unit 1 at each specified mileage interval to trigger the image acquisition unit 1 to collect track data. At the same time, the positioning synchronization mainboard 22 records the mileage position at the time of triggering and sends it to the image processing unit 4 through the switch 3. Since there is a cumulative error in calculating the moving mileage of the inspection vehicle 10 through the rotary encoder 21, the track line is divided into several mileage intervals and mileage labels are placed at the starting position of the interval. When the inspection vehicle 10 passes a certain mileage label, the label reader 23 reads the mileage information corresponding to the mileage label 24, updates the mileage data, and thus ensures the accuracy of the mileage data.

[0054] The switch 3 is located in the middle of the body 20 of the inspection vehicle 10, and is responsible for connecting each unit for data transmission. The switch 3 has no less than 4 Gigabit Ethernet interfaces for connecting the 3D inspection component (i.e., the image acquisition unit 1), the positioning synchronization system (i.e., the positioning synchronization unit 2) and the 3D inspection host (i.e., the image processing unit 4).

[0055] The track inspection system described in Example 1 of the present application mainly includes the following functions:

[0056] Data acquisition and storage: The track inspection system controls the 3D inspection components on the left and right sides to collect grayscale images and depth images of the track infrastructure at high speed and high resolution (to meet the vehicle's maximum speed requirement of 80km / h). The image acquisition unit 1 uses synchronous triggering technology, laser scanning imaging technology, multi-threading technology, etc. to achieve synchronous and equal mileage interval acquisition of left and right image information. The track inspection system stores the collected large-scale image data in a specially designed file format, and copies the long-term data to the disk array (i.e., storage unit 5) for long-term storage. The data file format contains basic track information, acquisition information, mileage information table, track grayscale image data, track depth image data, image sequence number, verification information, etc.

[0057] Data playback and historical data analysis: The track inspection system software enters the historical project playback function to play back and browse the collected images. The image playback software is divided into functional modules such as image browsing window, image switching and jump panel, and image operation panel. The image browsing window is mainly used to display the grayscale image of the track. With the mouse, the local area of ​​the image can be enlarged and reduced. The image switching and jump panel can realize the up and down switching of the image and jump by mileage; the image operation panel contains some sub-functions, which can realize the adjustment of the number of images displayed, image brightness adjustment, image contrast adjustment, image and 3D point cloud conversion, and image feature information annotation. The historical data analysis function can analyze the existing data and output corresponding reports to guide the further maintenance of the track infrastructure in the future.

[0058] Image recognition: A method based on deep learning of convolutional neural networks is used to automatically classify and identify surface defects of rails (wave wear, light band abnormalities, abrasions, peeling, fatigue cracks, high and low joints, etc.), fastener abnormalities (missing, displacement, fracture), sleepers (falling blocks, cracks), ballast (falling blocks, cracks), and induction plates (surface damage of induction plates, missing and loose anchor bolts) and calculate the size of damage. Specifically, a sample library (training set) is formed by collecting data on different types of damage, and the corresponding model is trained, so that the trained recognition model can be used to accurately identify various types of damage.

[0059] As attached Figure 6 As shown, the track inspection method based on the system described in Example 1 specifically includes the following steps:

[0060] When the inspection vehicle 10 is running, the rotary encoder 21 outputs pulse signals at equal angles (equal distances), the positioning synchronization mainboard 22 receives the pulse signals from the encoder 21, calculates the mileage at the current position in combination with the starting mileage, and sends a trigger signal to the image acquisition unit 1 at each specified mileage interval to trigger the image acquisition unit 1 to collect track data. At the same time, the positioning synchronization mainboard 22 records the mileage position at the time of triggering and sends it to the image processing unit 4 through the switch 3. The image processing unit 4 performs real-time or post-processing according to the track data and mileage position to obtain track damage identification data, and the storage unit 5 stores the original data or processed data output by the image processing unit 4.

[0061] The track inspection system described in Example 1 of the present application uses an engineering vehicle or electric passenger car as a platform, adopts a 3D camera to collect image data, combines grayscale images with three-dimensional point clouds for image recognition and three-dimensional measurement, and performs comprehensive three-dimensional detection of the track at the same time. The detection objects include rails, fasteners, sleepers, track plates and sensor plates. The specific embodiment of the present application provides a track (3D) inspection system and inspection vehicle for qualitative identification and quantitative measurement of visual defects of track infrastructure. It mainly uses line structured light imaging technology and adopts a 3D camera to collect the contour of the object to be measured illuminated by a laser, and obtains grayscale images and depth images of track infrastructure such as rails, fasteners, sleepers, and roadbeds along the track. Then, based on the mileage information and the collected grayscale images and depth images, a three-dimensional point cloud of the track is generated for subsequent detection and measurement.

[0062] Example 2

[0063] As attached Figure 7 As shown, an embodiment of the inspection vehicle 10 of the present application specifically includes: the track inspection system as described in Example 1. The rotary encoder is installed at the wheel of the inspection vehicle 10. The image acquisition unit 1 is installed at the rear of the vehicle body 20 of the inspection vehicle 10, as shown in the attached Figure 7 L in the middle indicates the direction of the vehicle head, that is, the operating direction of the inspection vehicle 10. The tag reader 23 is arranged near the image acquisition unit 1. The track inspection system also includes a mileage tag 24 arranged on the track line, and the tag reader 23 is used to obtain the mileage information corresponding to the mileage tag 24 after the inspection vehicle 10 passes through the mileage tag 24. The mileage tag 24 is arranged at the starting position of several mileage intervals divided by the track line. The positioning synchronization mainboard 22, the switch 3, the image processing unit 4 and the storage unit 5 are all arranged in the middle of the body 20 of the inspection vehicle 10.

[0064] The specific embodiment of the present application proposes a track (3D) inspection system and inspection vehicle, which have the characteristics of non-contact, fast speed, high precision, strong anti-interference, etc. The inspection system mainly uses line structured light imaging technology, collects the laser light strip image modulated by the height of the object surface, calculates the depth value and grayscale value of each point on the object surface on the laser line, and combines the trigger signal output by the inspection vehicle 10 (i.e., the detection platform) when it moves to obtain the mileage position of the laser line, and then obtains the complete track section profile information at the trigger position, and finally transmits it to the image processing unit 4 through Ethernet to obtain continuous track grayscale images and depth images. The inspection system automatically classifies and identifies rail surface defects including wave wear, light band anomalies, abrasions, peeling blocks, fatigue cracks, high and low joints, etc. through a deep learning method based on a convolutional neural network, and calculates the size of the damage. At the same time, a deep learning method based on a semantic neural network is used to automatically identify fastener anomalies such as missing fasteners and broken fasteners, and detailed information on sensor plate anomalies can be obtained.

[0065] In the description of this application, it should be noted that when an element is referred to as being "fixed to" or "disposed on" another element, it can be directly disposed on the other element or indirectly disposed on the other element. When an element is referred to as being "connected to" another element, it can be directly connected to the other element or indirectly connected to the other element.

[0066] It should be understood that the terms "center", "longitudinal", "lateral", "length", "width", "thickness", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside" and the like indicate orientations or positional relationships based on the orientations or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present application and simplifying the description, and do not indicate or imply that the referred device or element must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be understood as a limitation on the present application.

[0067] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the features. In the description of this application, "multiple" and "several" mean two or more, unless otherwise clearly and specifically defined.

[0068] It should be noted that the structures, proportions, sizes, etc. illustrated in the drawings of this specification are only used to match the contents disclosed in the specification for people familiar with this technology to understand and read, and are not used to limit the conditions under which this application can be implemented. Therefore, they have no substantive technical significance. Any structural modification, change in proportional relationship or adjustment of size should still fall within the scope of the technical content disclosed in this application without affecting the effects and purposes that can be achieved by this application.

[0069] By implementing the technical solutions of the track inspection system and the inspection vehicle described in the specific embodiments of this application, the following technical effects can be achieved:

[0070] (1) The track inspection system and inspection vehicle described in the specific embodiments of the present application establish three-dimensional track information and surface image information by acquiring track line grayscale images, depth images and mileage data, and use image recognition technology to accurately identify and locate abnormal track positions, which can solve the technical problem that the existing system only uses two-dimensional track grayscale images for track status inspection, and has a high false detection rate of rail surface damage;

[0071] (2) The rail inspection system and inspection vehicle described in the specific embodiments of the present application realize the 3D rail inspection for qualitative identification and quantitative measurement of visual defects of rail infrastructure. It can not only qualitatively identify the damage, but also quantitatively analyze the damage. It can also evaluate the development of the damage in combination with historical information and guide workers to perform maintenance.

[0072] (3) The track inspection system and inspection vehicle described in the specific embodiments of the present application are capable of performing 3D fine modeling and comprehensive three-dimensional inspection of the track surface. The inspection objects include track facilities such as rails, fasteners, sleepers, track plates and induction plates, and can realize the measurement and intuitive three-dimensional display of track diseases, foreign body detachment, etc.

[0073] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referenced to each other.

[0074] The above is only a preferred embodiment of the present application, and does not constitute any formal limitation to the present application. Although the present application has been disclosed as above with a preferred embodiment, it is not intended to limit the present application. Any technician familiar with the art can use the above disclosed methods and technical contents to make many possible changes and modifications to the technical solution of the present application, or modify it into an equivalent embodiment of equivalent changes, without departing from the spirit and technical solution of the present application. Therefore, any simple modification, equivalent replacement, equivalent change and modification made to the above embodiments according to the technical essence of the present application, which does not depart from the content of the technical solution of the present application, still falls within the scope of protection of the technical solution of the present application.

Claims

1. A track inspection system, characterized in that: include: An image acquisition unit (1), a positioning synchronization unit (2), a switch (3) and an image processing unit (4); the image acquisition unit (1), the switch (3) and the image processing unit (4) are connected in sequence, and the positioning synchronization unit (2) is connected to the image acquisition unit (1) and the switch (3) respectively; the positioning synchronization unit (2) sends a trigger signal to the image acquisition unit (1), triggering the image acquisition unit (1) to acquire 3D track data; the positioning synchronization unit (2) sends the mileage position data at the triggering moment to the image processing unit (4) through the switch (3); the 3D track data acquired by the image acquisition unit (1) is transmitted to the image processing unit (4) through the switch (3).

2. The track inspection system according to claim 1, characterized in that: The system further comprises a storage unit (5) connected to the image processing unit (4), and the storage unit (5) stores data output by the image processing unit (4).

3. The track inspection system according to claim 2, characterized in that: The positioning synchronization unit (2) comprises a rotary encoder (21), a positioning synchronization mainboard (22) and a tag reader (23); when the inspection vehicle (10) is running, the rotary encoder (21) outputs a pulse signal, the positioning synchronization mainboard (22) receives the pulse signal from the encoder (21), and sends a trigger signal to the image acquisition unit (1) at a set mileage interval to trigger the image acquisition unit (1) to collect track data; the positioning synchronization mainboard (22) sends the mileage position data at the time of triggering to the image processing unit (4) through the switch (3); the tag reader (23) is used to obtain the mileage information corresponding to a certain set position after the inspection vehicle (10) passes the position.

4. The track inspection system according to claim 3, characterized in that: The rotary encoder is mounted on the wheel axle of the inspection vehicle (10).

5. The track inspection system according to claim 3 or 4, characterized in that: The image acquisition unit (1) is installed at the rear of a vehicle body (20) of the inspection vehicle (10).

6. The track inspection system according to claim 5, characterized in that: The tag reader (23) is arranged close to the image acquisition unit (1).

7. The track inspection system according to claim 3, 4 or 6, characterized in that: The positioning synchronization mainboard (22), the switch (3), the image processing unit (4) and the storage unit (5) are all arranged in the middle of the vehicle body (20) of the inspection vehicle (10).

8. The track inspection system according to claim 7, characterized in that: The image processing unit (4) comprises a data acquisition and storage module (41), a data playback and historical data analysis module (42) and an image recognition module (43).

9. The track inspection system according to claim 3, 4, 6 or 8, characterized in that: The image acquisition unit (1) is used to acquire a laser light strip image modulated by the height of the object surface, and then calculate the grayscale value and depth value of the object surface, and finally obtain a grayscale image and a depth image of the object surface.

10. The track inspection system according to claim 9, characterized in that: The image acquisition unit (1) comprises a left 3D inspection component (11) and a right 3D inspection component (12) for respectively acquiring track data on the left and right sides. The left 3D inspection component (11) and the right 3D inspection component (12) shoot vertically downward and acquire track infrastructure data including left and right rails, fasteners, sleepers and track plates.

11. The track inspection system according to claim 10, characterized in that: The left 3D inspection component (11) and the right 3D inspection component (12) both comprise a line laser (101), a short-focus lens (102), a filter (103) and a 3D camera (104); the line laser (101) emits a line laser of a specific wavelength to the track surface, the line laser is highly modulated by the surface of the object to form a laser light strip, the laser light strip is collected by the 3D camera (104) after passing through the short-focus lens (102) and the filter (103) to form a laser light strip image, and then the grayscale value and depth value of the object surface are calculated, and finally a grayscale image and a depth image of the object surface are obtained.

12. The track inspection system according to claim 3, 4, 6, 8, 10 or 11, characterized in that: The system also includes a mileage tag (24) arranged on the track line, and the tag reader (23) is used to obtain mileage information corresponding to a mileage tag (24) after the inspection vehicle (10) passes by the mileage tag (24).

13. The track inspection system according to claim 12, characterized in that: The mileage labels (24) are arranged at the starting positions of a plurality of mileage intervals divided by the track line.

14. A patrol vehicle, characterized in that: include: A track inspection system as claimed in any one of claims 1 to 11.

Images