Rail transit vehicle bottom image acquisition method

Abstract

The invention discloses a rail transit vehicle bottom image acquisition method, an image is quickly acquired through a quick-scanning acquisition system, the quick-scanning acquisition system comprises a quick-scanning camera, an acquisition machine and an industrial personal computer, the quick-scanning camera is mounted on an inspection robot, and the acquisition machine is mounted on the inspection robot. The process of fast image acquisition of the fast-scanning acquisition system is as follows: the inspection robot forwards and continuously moves at the bottom of the rail transit vehicle based on an inspection path to inspect, in the inspection process, the industrial personal computer sends a fast-scanning acquisition instruction to the fast-scanning camera through the acquisition machine, a first folder path for fast-scanning image storage is generated in the acquisition machine, and a second folder path for fast-scanning image storage is generated in the acquisition machine; and the fast scanning camera collects a vehicle bottom image and stores the vehicle bottom image in the first folder path until the inspection robot reaches the end point of the forward inspection path, and fast image collection is completed. Through automatic inspection and high-precision image acquisition, the inspection efficiency and the data quality are remarkably improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of image acquisition, and in particular relates to a method for acquiring bottom images of rail transit vehicles. Background Art

[0002] As an important part of modern urban transportation, rail transit has attracted much attention for its safety and operational efficiency. With the continuous expansion of the rail transit network and the increase in the frequency of vehicle use, higher requirements are placed on the inspection and maintenance of rail transit vehicles. Especially for high-speed railway vehicles such as EMUs, due to their high running speed and heavy load, the safety status of bottom components such as bogies, braking systems, suspension devices, etc. is directly related to driving safety and passenger comfort. Therefore, the inspection and fault detection of the bottom of rail transit vehicles has become one of the key links to ensure the safe operation of vehicles.

[0003] Traditional rail transit vehicle underbody inspection and maintenance mainly rely on manual inspection. During manual inspection, maintenance personnel use handheld image acquisition devices to collect images of the underbody. This method has the following problems:

[0004] Low efficiency: Manual inspection requires professionals to enter the maintenance trench and check each component under the vehicle one by one. Due to the complex environment and small space under the vehicle, the work efficiency of the maintenance personnel is greatly limited. In addition, manual inspection often needs to be carried out during the train downtime, which further affects the inspection efficiency.

[0005] Insufficient data accuracy: Manual inspections rely on the experience and technical level of maintenance personnel and are easily affected by subjective factors, resulting in inaccurate inspection results. Some minor faults or potential problems may be overlooked, which in turn affects the safety of the vehicle.

[0006] Safety hazards: During the maintenance process, maintenance personnel need to work in a narrow trench, which poses certain safety risks. In addition, the bottom components of high-speed trains are complex, and some components have high temperatures during operation, which increases the difficulty and risk of maintenance. Summary of the invention

[0007] The purpose of the present invention is to solve the deficiencies of the above-mentioned background technology and provide a method for collecting bottom images of rail transit vehicles to improve collection efficiency and data quality.

[0008] The technical solution adopted by the present invention is: a method for acquiring images of the bottom of a rail transit vehicle, which performs rapid image acquisition through a quick scan acquisition system, wherein the quick scan acquisition system includes a quick scan camera, a collection machine and an industrial computer, and the quick scan camera is installed on a patrol robot. The process of rapid image acquisition by the quick scan acquisition system is as follows: the patrol robot performs forward continuous movement at the bottom of the rail transit vehicle based on an inspection path, and during the inspection process, the industrial computer sends a quick scan acquisition instruction to the quick scan camera through the collection machine, and a first folder path for storing quick scan images is generated in the collection machine. The quick scan camera collects images of the bottom of the vehicle and stores them in the first folder path until the patrol robot reaches the end of the forward patrol path to complete rapid image acquisition.

[0009] Furthermore, the quick-scan cameras are provided with a plurality of them. When the image is quickly collected, the port through which the industrial computer sends the scan and collection instructions to the collection machine is the same port, so that the plurality of quick-scan cameras collect images synchronously.

[0010] Furthermore, the image is accurately collected by a precision scanning collection system, the precision scanning collection system includes a precision scanning camera, the precision scanning camera is installed at the execution end of the inspection robot mechanical arm, and the process of the precision scanning collection system for accurately collecting images is as follows:

[0011] Step 1: After the image is quickly collected, the inspection robot performs inspections by intermittently moving in the reverse direction at the bottom of the rail transit vehicle based on the inspection path.

[0012] Step 2: When the inspection robot runs in reverse to the inspection station, the industrial computer calls the inspection trajectory of the robotic arm and sends a start-up command to the robotic arm.

[0013] Step 3: After receiving the start-up command, the robot arm sends a collection command to the collection machine through the industrial computer, and the collection machine generates a second folder path for storing the precision scanned image according to the collection command;

[0014] Step 4, when the robot arm reaches the inspection point, the gain information is sent to the acquisition machine through the industrial computer, and the acquisition machine controls the precision scanning camera to acquire the bottom image of the vehicle based on the gain information and stores it in the second folder path;

[0015] Step 5: After the image acquisition is completed, the industrial computer sends the acquisition feedback information to the robot arm, and the robot arm moves to the next inspection point and repeats steps 4-5 until the image acquisition of all inspection points on the inspection track is completed;

[0016] Step 6: The inspection robot continues to run in the reverse direction to the next inspection station, and repeats steps 2-6 until image acquisition of all inspection stations on the inspection path is completed.

[0017] Furthermore, the precision scanning cameras are provided in plurality, and the industrial control computer and the acquisition machine are used to control the plurality of precision scanning cameras to work asynchronously to accurately acquire images.

[0018] Furthermore, there are multiple first forwarding ports connecting the industrial computer and the robotic arm, and multiple second forwarding ports connecting the industrial computer and the acquisition machine. For different precision scanning cameras, multiple precision scanning cameras can work asynchronously to accurately acquire images through the multiple first forwarding ports connected between the industrial computer and the robotic arm and the multiple second forwarding ports connected between the industrial computer and the acquisition machine.

[0019] Furthermore, the network ports of the robotic arm, industrial computer, acquisition machine, and precision scanning camera are respectively allocated different network IPs to realize the transmission of instructions or data.

[0020] Furthermore, the industrial computer and the robotic arm communicate via TCP protocol.

[0021] Furthermore, the industrial computer and the collection machine communicate with each other via UDP protocol.

[0022] Furthermore, during the image acquisition process, the acquisition machine uploads the acquired images to the algorithm server in real time.

[0023] The beneficial effects of the present invention are:

[0024] The present invention uses an automated inspection and high-precision image acquisition (a combination of fast scanning and precise acquisition) system. The fast scanning image acquisition module can quickly acquire 2D and 3D images of the bottom of the EMU when the detection robot is running in the forward direction; while the precise scanning image acquisition module performs a more accurate scan when the robot runs in the reverse direction to the bogie position, thereby realizing accurate and efficient image acquisition in complex environments, ensuring real-time data collection, uploading and processing, significantly improving high-speed rail inspection efficiency and data quality, and reducing labor costs and error rates.

[0025] The present invention realizes efficient data transmission and processing, improves system stability and maintainability, enhances the safety and reliability of high-speed rail inspection, and has broad application prospects and market potential by designing the network topology, communication protocol and automation process of the robotic arm, industrial computer, acquisition machine and precision scanning camera, including IP address allocation among components and optimization of communication protocol.

[0026] When accurately capturing images, the present invention adjusts the parameters of the precision scanning camera by sending gain information through a mechanical arm, thereby ensuring the consistency and integrity of image capture and avoiding data quality problems caused by environmental changes. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 It is a schematic diagram of the quick scan acquisition system of the present invention.

[0028] Figure 2 It is a schematic diagram of the composition of the quick scan camera of the present invention.

[0029] Figure 3 The present invention is a flow chart for accurately capturing images.

[0030] Figure 4 Schematic diagram of the distribution of network IP of the precision scanning collection system of the present invention. DETAILED DESCRIPTION

[0031] The specific embodiments of the present invention are further described below in conjunction with the accompanying drawings. It should be noted that the description of these embodiments is used to help understand the present invention, but does not constitute a limitation of the present invention. In addition, the technical features involved in the various embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0032] The present invention provides a method for collecting images of the bottom of a rail transit vehicle. First, a fast scanning collection system is used to quickly collect images. Figure 1 As shown, the quick scan acquisition system includes a quick scan camera, an acquisition machine and an industrial computer. The quick scan camera is installed on the inspection robot. The process of quick scan acquisition system for quick image acquisition is: the inspection robot performs inspection in the bottom of the rail transit vehicle based on the inspection path by continuous forward movement (i.e., at a constant speed). During the inspection, the industrial computer sends a quick scan acquisition instruction to the quick scan camera through the acquisition machine, and the acquisition machine generates a first folder path for quick scan image storage. The quick scan camera acquires the image of the bottom of the vehicle and stores it in the first folder path until the inspection robot reaches the end of the forward inspection path to complete the quick image acquisition.

[0033] In the above scheme, the quick scan camera installed on the inspection robot can quickly collect the bottom image of the rail transit vehicle. The quick scan camera is composed of an area array camera and a line array camera. Figure 2As shown, the area array transmission and the area array reception use triangulation to obtain 3D point cloud image information, and the line array transmission and the line array reception use line laser to obtain 2D black and white image information. The common area of ​​the area array and the line array acquisition is the acquisition depth of field of the quick scan camera. Within the depth of field, 2D and 3D aligned images in three formats of .jpg, .png, and -x.jpg can be obtained. The quick scan acquisition system of the present invention realizes high-precision 2D and 3D image acquisition of the bottom of the EMU. The high-speed line array camera can quickly capture the detailed image of the bottom of the vehicle, and the three-dimensional area array camera can generate a three-dimensional model of the bottom component to provide more comprehensive detection data. Multiple quick scan cameras can be set. When the image is quickly acquired, the port of the quick scan acquisition instruction for multiple quick scan cameras sent by the industrial control computer to the acquisition machine is the same port, so that multiple quick scan cameras synchronously acquire images. During the image acquisition process, the acquisition machine uploads the acquired image to the algorithm server in real time. The algorithm server is connected to the detection workstation, the edge inference machine and the handheld terminal, thereby facilitating image processing or management at the workstation or handheld terminal. The present invention has an automatic data uploading function. The detected images and fault information can be transmitted to the detection workstation and edge inference machine in real time for processing and storage. The handheld mobile terminal can also help maintenance personnel read and upload maintenance results anytime and anywhere to realize information management.

[0034] In the above scheme, after the rapid image acquisition is completed, the image is accurately acquired through the fine scanning acquisition system. The fine scanning acquisition system includes a fine scanning camera, which is installed at the execution end of the mechanical arm of the inspection robot. The fine scanning camera collects color two-dimensional images and point cloud images by emitting and receiving structured light. Through the taught trajectory, the fine scanning camera can reach areas that the fast scanning camera cannot cover or cover poorly, and perform precise scanning.

[0035] like Figure 3 As shown, the process of the precise image acquisition system is as follows:

[0036] Step 1: After the image is quickly collected, the inspection robot performs inspection based on the inspection path by intermittently moving in the reverse direction at the bottom of the rail transit vehicle (i.e., it stops at each station, collects images, and then moves to the next station).

[0037] Step 2: When the inspection robot runs in reverse to the inspection station, the industrial computer calls the inspection trajectory of the robotic arm and sends a start-up command to the robotic arm.

[0038] Step 3: After receiving the start-up command, the robot arm sends a collection command to the collection machine through the industrial computer, and the collection machine generates a second folder path for storing the precision scanned image according to the collection command;

[0039] Step 4: When the robot arm reaches the inspection point, the industrial computer sends gain information to the acquisition machine. The acquisition machine controls the precision scanning camera based on the gain information to acquire the bottom image of the vehicle and stores it in the second folder path. During the image acquisition process, the acquisition machine uploads the acquired image to the algorithm server in real time.

[0040] Step 5: After the image acquisition is completed, the industrial computer sends the acquisition feedback information to the robot arm, and the robot arm moves to the next inspection point and repeats steps 4-5 until the image acquisition of all inspection points on the inspection track is completed;

[0041] Step 6: The inspection robot continues to run in the reverse direction to the next inspection station, and repeats steps 2-6 until image acquisition of all inspection stations on the inspection path is completed.

[0042] In the above scheme, the precision scanning camera is provided with multiple, specifically two, respectively, a left precision scanning camera and a right precision scanning camera. The two precision scanning cameras are respectively arranged at the ends of the two mechanical arms, and the two precision scanning cameras are controlled by the industrial control computer and the acquisition machine to work asynchronously and accurately collect images. The industrial control computer is provided with multiple first forwarding ports connected to the mechanical arm, and multiple second forwarding ports connected to the acquisition machine are provided, so as to distinguish different precision scanning cameras. For different precision scanning cameras, multiple precision scanning cameras can work asynchronously and accurately collect images through multiple first forwarding ports connected between the industrial control computer and the mechanical arm and multiple second forwarding ports connected between the industrial control computer and the acquisition machine. Specifically, for the left precision scanning camera, the industrial control computer establishes communication with the mechanical arm through the forwarding port 7001, sends instructions to the acquisition machine through the acquisition sending port 8001, and receives instructions from the acquisition machine through the acquisition receiving port 8000; for the right precision scanning camera, the industrial control computer establishes communication with the mechanical arm through the forwarding port 7101, sends instructions to the acquisition machine through the acquisition sending port 8101, and receives instructions from the acquisition machine through the acquisition receiving port 8100.

[0043] In the above scheme, the network ports of the robot arm, industrial computer, acquisition machine, and precision scanning camera are respectively assigned different network IPs to realize the transmission of instructions or data. The specific allocation addresses are as follows: Figure 4 As shown. The industrial computer and the manipulator communicate via the TCP protocol, and the industrial computer and the collector communicate via the UDP protocol. The present invention achieves efficient data transmission and processing, improves system stability and maintainability, enhances the safety and reliability of high-speed rail inspection, and has broad application prospects and market potential by designing the network topology, communication protocol and automation process of the manipulator, industrial computer, collector and precision scanning camera, including IP address allocation and communication protocol optimization between components.

[0044] Example 1

[0045] 1. Setting of port numbers of each device in the quick scan acquisition system:

[0046] (1) The industrial computer sends a collection instruction to the collection machine protocol forwarding program: the industrial computer sends the port number: 6100; the collection machine protocol forwarding program receives the port number: 6100.

[0047] (2) The acquisition machine protocol forwarding program sends acquisition instructions to the quick scan camera acquisition program: the protocol forwarding program sends port number: 6301; the quick scan camera acquisition program receives port number: 6301.

[0048] (3) The quick scan camera acquisition program feeds back instructions to the protocol forwarding program: the quick scan acquisition program sends port number: 6300; the protocol forwarding program receives port number: 6300.

[0049] (4) The protocol forwarding program gives feedback instructions to the industrial computer: the protocol forwarding program of the data acquisition machine sends port number: 6101; the industrial computer receives port number: 6101.

[0050] (5) The quick scan acquisition program sends instructions to the quick scan scheduling program: the quick scan acquisition program sends port number: 20003; the quick scan scheduling program receives port number: 20003.

[0051] (6) The quick scan acquisition program sends an upload instruction to the quick scan transmission program: the quick scan acquisition program sends port number: 62000; the quick scan transmission program receives port number: 62000.

[0052] (7) The quick scan transmission program sends a receiving instruction to the server (Linux) quick scan receiving program: the quick scan transmission program sends the port number: 42001; the server (Linux) quick scan receiving program receives the port number: 42001.

[0053] (8) The industrial computer sends an end-collection instruction to the collector's protocol forwarding program: the industrial computer sends the port number: 6100; the collector's protocol forwarding program receives the port number: 6100.

[0054] (9) The protocol forwarding program of the acquisition machine sends an end acquisition instruction to the quick scan acquisition program: the protocol forwarding program sends the port number: 6301; the quick scan acquisition program receives the port number: 6301.

[0055] (10) The quick scan acquisition program feeds back the acquisition end instruction to the protocol forwarding program: the quick scan acquisition program sends the port number: 6300; the protocol forwarding program receives the port number: 6300.

[0056] (11) The protocol forwarding program feeds back the acquisition end instruction to the industrial computer: the acquisition machine protocol forwarding program sends the port number: 6101; the industrial computer receives the port number: 6101.

[0057] 2. The steps of quick scan image acquisition and upload process are as follows:

[0058] (1) After the system is turned on, the industrial computer is in working state for a long time, and other devices are in powered-off state.

[0059] (2) After the system is powered on, power is supplied to the quick scan camera and the data collector at the same time.

[0060] (3) At this time, the industrial computer sends a status query command: CS&STATE to the acquisition program: 61000 on the acquisition machine 192.168.3.34 through 192.168.3.35:6100, and the acquisition machine has no return value.

[0061] (4) After about 300 seconds, the data collector starts up successfully and all data collection and upload related software are started.

[0062] (5) At this time, the quick scan acquisition software begins to initialize the quick scan camera.

[0063] The control machine sends a status query command: CS&STATE to the acquisition program: 61000 on the acquisition machine 192.168.3.34 through 192.168.3.35:6100. The acquisition program on the acquisition machine queries port 41001 and returns: DISABLE.

[0064] (6) After about 60 seconds, the acquisition machine's quick scan acquisition software completes the initialization of the four cameras one by one, and displays that the camera status is ready.

[0065] The control machine sends a status query command: CS&STATE to the acquisition program: 61000 on the acquisition machine 192.168.3.34 through 192.168.3.35:6100. The acquisition program on the acquisition machine queries port 41001 and returns: ENABLE.

[0066] (7) Now you can perform the quick scan camera collection and upload work normally

[0067] (8) The inspection robot starts inspection, and the industrial computer sends a quick scan collection command to the collection machine:

[0068] CS&OPEN&CurrentImagePath (CurrentImagePath is the current image storage path)

[0069] example:

[0070] CS&OPEN&E:\RawImageDisk\FastScan_raw\J11\20220711\2215\C RH2A2411\202207112215CRH2A2411

[0071] At this time, the folder path is generated in the collector:

[0072] E:\RawImageDisk\FastScan_raw\ID track number\year month day\hour and minute\car model number\task ID (year month day hour and minute car model number).

[0073] example:

[0074] E:\RawImageDisk\FastScan_raw\J11\20220711\2215\CRH2A2411\202207112215CRH2A2411.

[0075] The quick scan camera is in the collection state.

[0076] (9) After the quick scan camera collects the status, the acquisition machine feeds back a signal to the industrial computer:

[0077] CS&COMPLETE

[0078] Tell the industrial computer that the quick scan camera is working properly.

[0079] (10) During the inspection process of the inspection robot, the path where the collected images are saved on the acquisition machine is:

[0080] E:\RawImageDisk\FastScan_raw\ID track number\year month day\hour and minute\car model number\task ID (year month day hour and minute car model number).

[0081] The original image generated in the acquisition machine: B1-original image name-x.jpg (2D)

[0082] B1-Original picture name.jpg(3D)

[0083] B1-Original picture name.png

[0084] B2-Original picture name-x.jpg(2D)

[0085] B2-Original picture name.jpg(3D)

[0086] B2-Original picture name.png

[0087] (11) During the inspection process, the transmission program of the acquisition machine transmits the original image to the server (Linux) in real time. The storage path on the server is:

[0088] media\bjtds\RawImageDisk\FastScan_raw\ID track number\year-month-day\hour-minute\car model number\task ID (year-month-day-hour-minute car model number).

[0089] The original image is generated on the server: B1-original image name-x.jpg(2D)

[0090] B1-Original picture name.jpg(3D)

[0091] B1-Original picture name.png(3D)

[0092] B2-Original picture name-x.jpg(2D)

[0093] B2-Original picture name.jpg(3D)

[0094] B2-Original picture name.png(3D)

[0095] (12) After the quick scan camera completes the acquisition, the industrial computer sends the acquisition end command to the acquisition machine: CS&FINISH;

[0096] After receiving the command to end the acquisition, the acquisition program will feedback the command to the industrial computer: CS&CLOSE.

[0097] Example 2

[0098] 1. The network IP allocation of each device in the precision scanning collection system is as follows:

[0099] (1) Right robotic arm network port IP: 192.168.3.120

[0100] (2) Left robotic arm network port IP: 192.168.3.121

[0101] (3) Industrial computer control network port IP: 192.168.3.35

[0102] (4) Industrial computer transmission network port IP: 192.168.11.35

[0103] (5) Collector control network port IP: 192.168.3.34

[0104] (6) Collector transmission network port IP: 192.168.11.34

[0105] (7) The IP address of the collection machine’s right precision scanning collection network port is: 192.168.1.10

[0106] (8) The left precision scanning acquisition network port IP of the data collector is: 192.168.2.10

[0107] (9) Scan the CPE wireless network IP: 192.168.1.11

[0108] (10) Scan the CPE wireless network IP from the left: 192.168.2.11

[0109] (11) Scan the network IP: 192.168.1.120

[0110] (12) Scan the left network IP: 192.168.2.120

[0111] (13) Algorithm server network IP: 192.168.0.201

[0112] 2. TCP / UDP communication process and port settings between the precision scanning camera, robotic arm, and industrial computer:

[0113] Left precision scanning camera:

[0114] (1) Left precision scan camera number: T1.

[0115] (2) The robot arm TCP client connects to the industrial computer TCP service port: 7001.

[0116] (3) The industrial computer binds to UDP port: 8001 and sends collection instructions to the collection machine UDP port: 8001.

[0117] (4) The acquisition machine controls the left precision scanning camera to complete image acquisition, and the image is generated normally.

[0118] (5) The data collector binds to UDP port: 8000 and sends a completion signal to the industrial computer UDP port: 8000.

[0119] (6) The acquisition machine transmits the acquired images to the algorithm server.

[0120] Right precision scanning camera:

[0121] (1) Right precision scanning camera number: T2.

[0122] (2) The robot arm TCP client connects to the industrial computer TCP service port: 7101.

[0123] (3) The industrial computer binds to UDP port: 8101 and sends collection instructions to the collection machine UDP port: 8101.

[0124] (4) The acquisition machine controls the left precision scanning camera to complete image acquisition, and the image is generated normally.

[0125] (5) The data collector binds to UDP port: 8100 and sends a completion signal to the industrial computer UDP port: 8100.

[0126] (6) The acquisition machine transmits the acquired images to the algorithm server.

[0127] 3. Introduction to the storage path and naming rules of the original images of the precision scanned images.

[0128] The storage path of the original image in the acquisition machine (Win10):

[0129] C:\RawImageDisk\PreScan_raw\J11\20240423\1143\CR400AF2200\202404231143C R400AF2200.

[0130] The storage path of the original image in the server (linux):

[0131] media\bjtds\RawImageDisk\PreScan_raw\J11\20240423\1143\CR400AF2200\202404 231143CR400AF2200.

[0132]

[0133]

[0134] 4. The steps of accurate image acquisition and uploading process are:

[0135] (1) The inspection robot starts inspection, completes the forward quick scan and collection, and runs in the reverse direction to reach the designated workstation. The industrial computer calls the robot arm trajectory program through the PLC.

[0136] (2) The robot arm acts as a client to connect to the industrial computer service port (7001 / 7101) to establish TCP communication, which is hereinafter referred to as the industrial computer forwarding port.

[0137] (3) The robotic arm sends the image acquisition start command through the industrial computer forwarding port: START&!

[0138] (4) After the industrial computer receives the collection start command from the industrial computer forwarding port, it parses and synthesizes the collection preparation command: START&track number&year-month-day&hour-minute&vehicle model number&task number (year-month-day-hour-minute vehicle model number).

[0139] example:

[0140] START&J11&20240423&1143&CR400AF2200&202404231143CR 400AF2200.

[0141] (5) The industrial computer binds UDP port 8001 / 8101 and sends a preparation acquisition instruction to the acquisition machine UDP port: 8001 / 8101, which is collectively referred to as the industrial computer acquisition sending port. At this time, the folder path generated in the acquisition machine is: C:\RawImageDisk\PreScan_raw\track number\year-month-day\hour-minute\car model number\task number\.

[0142] example:

[0143] C:\RawImageDisk\PreScan_raw\J11\20240423\1143\CR400AF2200\202404231143C R400AF2200.

[0144] At this time, the precision scanning camera is in the waiting state.

[0145] (6) When the robot arrives at the inspection point, it sends the gain information to the industrial computer through the industrial computer forwarding port: CHANGE & GAIN & gain value! .

[0146] Example: CHANGE&GAIN&20!

[0147] (7) The robot arm sends the image acquisition information to the industrial computer through the industrial computer forwarding port: PHOTO&component number-first-level component number-second-level component number-inspection point number! .

[0148] Example: PHOTO&01-ZDZZ-05-08! This means that the inspection point No. 8 is the No. 1 trailer brake device No. 05.

[0149] (8) The industrial computer sends a gain modification command to the acquisition machine through the industrial computer acquisition transmission port: CHANGE&GAIN&gain value&camera number.

[0150] Example: CHANGE&GAIN&20&T1

[0151] (9) The industrial computer sends the image acquisition command to the acquisition machine through the industrial computer acquisition and transmission port: PHOTO&car number, bogie number&original image name.

[0152] The original image name consists of: T1-year-month-day-vehicle model number-carriage number-bogie number-axle number-component number.

[0153] Example of first-level component name-second-level component number-inspection point number:

[0154] PHOTO&01A&T1-20211106-CRH2A2411-01-A-01-01-ZDZZ-05-08.

[0155] (10) At this time, the precision scanning camera is triggered to take a photo and the collected image data is stored in the folder generated in (1): N-T1-year-month-day-vehicle model number-carriage number-bogie number-axle number-component position number-first-level component name-second-level component number-inspection point number.

[0156] Example: N-T1-20211106-CRH2A2411-01-A-01-01-ZDZZ-05-08.bmp N-T1-20211106-CRH2A2411-01-A-01-01-ZDZZ-05-08.tiff

[0157] N-T1-20211106-CRH2A2411-01-A-01-01-ZDZZ-05-08.jpg

[0158] N-T1-20211106-CRH2A2411-01-A-01-01-ZDZZ-05-08.xyz

[0159] N-T1-20211106-CRH2A2411-01-A-01-01-ZDZZ-05-08.png.

[0160] (11) After the image is generated normally, the acquisition machine sends the acquisition completion instruction: PHOTO&COMPLETE to the industrial computer UDP port 8000 / 8100 (hereinafter referred to as the industrial computer acquisition receiving port) through UDP port 8000 / 8100.

[0161] The image is uploaded to the algorithm server in real time.

[0162] (12) The industrial computer receives the collection completion instruction through the industrial computer collection receiving port, analyzes and synthesizes, and collects feedback information!

[0163] (13) The industrial computer sends the collected feedback information to the robotic arm through the industrial computer forwarding port!

[0164] At this point, the robot arm moves to the next inspection point and repeats steps (6)-(13).

[0165] (14) When all inspection points of the workstation are collected, the robot arm returns to its initial posture, and the robot arm sends the collection end command PHOTO&FINISH! to the industrial computer through the industrial computer forwarding port.

[0166] (15) The industrial computer sends the acquisition end instruction PHOTO&FINISH to the acquisition machine through the industrial computer acquisition port.

[0167] (16) The inspection robot continues to run in reverse to the next station and repeats steps (1)-(16)

[0168] (17) Wait until all inspection stations have completed data collection and exit the one-key inspection function.

[0169] The above are only specific embodiments of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily thought of by a person skilled in the art within the technical scope disclosed by the present invention should be included in the protection scope of the present invention. The contents not described in detail in this specification belong to the prior art known to professional and technical personnel in the field.

Claims

1. A method for collecting images of the bottom of a rail transit vehicle. Features: The quick-scan acquisition system is used for quick image acquisition. The quick-scan acquisition system includes a quick-scan camera, an acquisition machine and an industrial computer. The quick-scan camera is installed on the inspection robot. The process of quick-scan acquisition by the quick-scan acquisition system is as follows: the inspection robot performs inspections in a forward continuous motion at the bottom of the rail transit vehicle based on the inspection path. During the inspection, the industrial computer sends a quick-scan acquisition instruction to the quick-scan camera through the acquisition machine. The acquisition machine generates a first folder path for storing quick-scan images. The quick-scan camera acquires images of the bottom of the vehicle and stores them in the first folder path until the inspection robot reaches the end of the forward inspection path to complete quick image acquisition.

2. The method for collecting bottom images of rail transit vehicles according to claim 1, Features: The quick-scan cameras are provided with a plurality of them. When the image is quickly collected, the port through which the industrial computer sends the scan and collection instructions to the collection machine is the same port, so that the plurality of quick-scan cameras collect images synchronously.

3. The rail transit vehicle bottom image acquisition method according to claim 1, Features: The image is also accurately collected by a precision scanning collection system, the precision scanning collection system includes a precision scanning camera, and the precision scanning camera is installed at the execution end of the inspection robot mechanical arm. The process of the precision scanning collection system for accurately collecting images is as follows: Step 1: After the image is quickly collected, the inspection robot performs inspections by intermittently moving in the reverse direction at the bottom of the rail transit vehicle based on the inspection path; Step 2: When the inspection robot runs in the reverse direction to the inspection station, the industrial computer calls the inspection trajectory of the robotic arm and sends a start-up command to the robotic arm; Step 3: After receiving the start-up command, the robot arm sends a collection command to the collection machine through the industrial computer, and the collection machine generates a second folder path for storing the precision scanned image according to the collection command; Step 4, when the robot arm reaches the inspection point, the gain information is sent to the acquisition machine through the industrial computer, and the acquisition machine controls the precision scanning camera to acquire the bottom image of the vehicle based on the gain information and stores it in the second folder path; Step 5: After the image acquisition is completed, the industrial computer sends the acquisition feedback information to the robot arm, and the robot arm moves to the next inspection point and repeats steps 4-5 until the image acquisition of all inspection points on the inspection track is completed; Step 6: The inspection robot continues to run in the reverse direction to the next inspection station, and repeats steps 2-6 until image acquisition of all inspection stations on the inspection path is completed.

4. The method for collecting bottom images of rail transit vehicles according to claim 3, Features: The precision scanning cameras are provided in plurality, and the industrial control computer and the acquisition machine are used to control the precision scanning cameras to work asynchronously and accurately acquire images.

5. The method for collecting bottom images of rail transit vehicles according to claim 4, Features: There are multiple first forwarding ports connecting the industrial computer and the robotic arm, and multiple second forwarding ports connecting the industrial computer and the acquisition machine. For different precision scanning cameras, multiple precision scanning cameras can work asynchronously and accurately acquire images through the multiple first forwarding ports connected between the industrial computer and the robotic arm and the multiple second forwarding ports connected between the industrial computer and the acquisition machine.

6. The method for collecting bottom images of rail transit vehicles according to claim 3, Features: The network ports of the robotic arm, industrial computer, acquisition machine, and precision scanning camera are respectively assigned different network IPs to realize the transmission of instructions or data.

7. The method for collecting bottom images of rail transit vehicles according to claim 1 or 3, Features: The industrial computer and the robotic arm communicate with each other via the TCP protocol.

8. The method for collecting bottom images of rail transit vehicles according to claim 1 or 3, Features: The industrial computer and the collection machine communicate via the UDP protocol.

9. The method for collecting bottom images of rail transit vehicles according to claim 1 or 3, Features: During the image acquisition process, the acquisition machine uploads the acquired images to the algorithm server in real time.

Images