Pantograph slide plate image acquisition method and system

Abstract

This invention discloses a method and system for acquiring pantograph contactor images, belonging to the field of railway power supply contact network detection technology. It uses a photoelectric sensor signal as a trigger. When the photoelectric sensor module detects an approaching train, it controls the pantograph imaging module to take a picture based on a system-preset speed and the radar's measured speed. After the train passes, the image's stretching / compression ratio is calculated based on the radar's final detected speed and the system-preset speed. The identified pantograph contactor image is then stretched, compressed, and segmented to obtain the final pantograph contactor image. By introducing a photoelectric sensor, both the sensor and radar can identify approaching trains. Furthermore, by combining the imaging process based on real-time radar speed measurement and a system-preset speed, even if a train is missed by the radar, the entire image acquisition process can be completed based on the photoelectric sensor's identification and the system-preset speed, avoiding the problem of missed trains by the radar.

Description

Technical Field

[0001] This invention relates to the field of railway power supply contact network detection technology, specifically to a method and system for acquiring pantograph sliding plate images. Background Technology

[0002] A pantograph is an electrical device used by electric traction locomotives (such as high-speed trains, bullet trains, electric locomotives, and subways) to obtain electrical energy from the overhead contact line. The pantograph is mounted on the roof of the train, and when raised, its top sliding plate makes close contact with the contact wire, allowing high-voltage electricity (usually 25kV AC) to be introduced from the power grid into the train. The current then flows through the pantograph into the train's traction system, driving the motor. To ensure the safe operation of the railway power supply contact network, it is necessary to monitor the pantograph's sliding plate, which requires acquiring images of the pantograph's sliding plate.

[0003] The traditional method for acquiring pantograph images involves using a speed radar to detect passing vehicles. When a vehicle is detected, a high-frame-rate monochrome camera continuously captures images of the entire vehicle to obtain the license plate number. Two shooting schemes are used: a high-definition line-scan camera or a high-resolution area-scan camera. However, this traditional method has the following problems:

[0004] 1. Train arrival monitoring: Various manufacturers use common speed measuring radars for train reception. However, for high-speed trains, the radar has a monitoring lag problem, which leads to the train's front pantograph sliding plate and train number being missed.

[0005] 2. For conventional locomotives with blue bodies and red license plates, images captured by black and white cameras will cause the license plate number and the background of the car body to blend into one, making it impossible for inspectors to distinguish the license plate number with the naked eye during platform analysis.

[0006] 3. The vehicle license plate images are continuously captured by a high frame rate camera, resulting in a large number of images, a large amount of data, and high bandwidth consumption during transmission. In densely trafficked road sections, the transmission time is long, which affects the system's next vehicle reception.

[0007] 4. Currently, pantograph imaging mainly employs two imaging schemes: high-definition line scan cameras or high-resolution high-definition area scan cameras. 1) Using high-definition line scan cameras presents two problems: a. The pantograph lacks positioning, resulting in the pantograph slide being segmented by two adjacent images, affecting platform data analysis; b. For high-speed trains traveling at 350 km / h, the triggering frequency of high-definition line scan cameras is limited, leading to image compression issues. 2) Using high-resolution high-definition area scan cameras, various manufacturers use photoelectric sensors for positioning, which can match high-speed trains traveling at 350 km / h. However, there is a problem where the photoelectric sensors may fail to trigger or not trigger at a train speed of 350 km / h, resulting in missed or lost images of the pantograph slide. Summary of the Invention

[0008] In view of this, the purpose of the present invention is to provide a method and system for acquiring pantograph sliding plate images, so as to solve the problem of missed trains by existing high-speed train radar.

[0009] To achieve the above objectives, the present invention provides a method for acquiring images of a pantograph sliding plate. Specifically, the method introduces a photoelectric sensor to form a dual-switch vehicle detection mechanism with radar. The photoelectric sensor is oriented towards the direction of oncoming vehicles. The method includes the following steps:

[0010] 1) Triggered by the photoelectric sensor signal, when the photoelectric sensor module detects an approaching vehicle, it controls the pantograph imaging module to take pictures based on the system's preset speed and the radar's measured speed.

[0011] 2) After the vehicle passes, the image stretching and compression ratio is calculated based on the final speed detected by the radar and the preset speed of the system. The identified pantograph sliding plate image is then stretched, compressed, and segmented to obtain the final pantograph sliding plate image.

[0012] The beneficial effects are as follows: By introducing a photoelectric sensor, the method of this invention forms a dual-switch vehicle detection mechanism with radar. Both the photoelectric sensor and radar can identify whether a vehicle is approaching. Furthermore, by combining the real-time radar speed measurement and the system preset speed to control the image capture process, even if a vehicle is missed by the radar, the complete image acquisition process can still be completed based on the photoelectric sensor's identification and the system preset speed, avoiding the problem of radar-missed vehicles. Moreover, this invention also considers the difference between the system preset speed and the final radar detection speed (i.e., the radar's measured speed), which may lead to a difference in scale between the acquired image and the actual image. Therefore, this invention further performs stretching, compression, and segmentation operations on the pantograph contactor image based on the stretching / compression ratio obtained from the final radar detection speed and the system preset speed, resulting in a final pantograph contactor image that matches the actual situation.

[0013] Further, in step 1), combining the system's preset speed and the radar's measured speed, controlling the pantograph's imaging module to take pictures includes:

[0014] When the photoelectric sensor module detects an approaching vehicle, if the radar has also detected the vehicle in advance, the radar speed measurement will be used to control the pantograph camera module to take a picture.

[0015] When the photoelectric sensor module detects an approaching vehicle, if the radar has not detected the vehicle beforehand, the pantograph camera module will take a picture using the system's preset speed control.

[0016] The method of this invention introduces photoelectric sensors and a system preset speed to be used in cases where the radar misses a train. If the radar does not detect the passing train in advance, it may lead to missed detection of the positions of the train ahead. To avoid this situation, this invention uses a system preset speed to control the pantograph imaging module to take pictures. However, if the radar can detect the passing train in advance, the pantograph imaging module is controlled to take pictures according to the actual speed measured by the radar. The resulting image is an image that matches the actual situation. Therefore, when the radar also detects the passing train in advance, the method of this invention controls the pantograph imaging module to take pictures by radar speed measurement, ensuring that the acquired image matches the actual scale and simplifying the subsequent stretching and compression process.

[0017] Furthermore, in step 1), the system preset speed is a speed configured for different time periods.

[0018] The method of this invention takes into account that since the types and speeds of vehicles passing through at fixed locations and at fixed times are basically the same, the preset speed of the system is set based on different time periods, so that the preset speed of the system is closer to the actual situation at that time. Therefore, when the camera with the current time period is used to take pictures, although the acquired image may deviate from the actual situation, the deviation is minimized as much as possible.

[0019] Further, in step 2), the calculation of the image stretching and compression ratio based on the final speed detected by the radar and the preset speed of the system includes: if the radar also detected the passing vehicle in advance, the stretching and compression ratio is 1; if the radar did not detect the process in advance, the stretching and compression ratio is the ratio of the radar's final detected speed to the preset speed of the system.

[0020] The method of this invention uses a photoelectric sensor to detect when a high-speed train passes by, and combines the system's preset speed and the radar's final measured speed to dynamically calculate the image stretching and compression ratio, and perform correction processing on the original image to ensure that the final image is displayed normally.

[0021] Further, in step 2), stretching, compression, and segmentation operations are performed on the identified pantograph slide image to obtain the final pantograph slide image, including: image enhancement of the image obtained by the pantograph imaging module, detection of a single image containing the pantograph slide area, and determination of the pantograph slide integrity of the single image.

[0022] If the completeness meets the threshold and the radar has also detected the passing vehicle in advance, then the single image is determined as the final pantograph skateboard image.

[0023] If the integrity does not meet the threshold, and the radar has detected the passing vehicle in advance, then the previous or next image of the single image is stitched together depending on whether the pantograph sliding plate area is located at the left head position or the right tail position of the single image. The pantograph sliding plate area is located in the stitched image to obtain the geometric center position of the pantograph sliding plate. Based on the resolution size of the single image, the stitched image is segmented with the geometric center position as the image center position to obtain the final pantograph sliding plate image.

[0024] If the radar does not detect the vehicle in advance, then based on the single image, take two images in front and two images behind, for a total of five images, and stitch them together. Then, perform a stretching and compression operation on the stitched image according to the stretching and compression ratio. Using the geometric center of the stretched and compressed pantograph slide in the image as the origin and the resolution size of the single image as the reference, perform a segmentation operation on the stretched and compressed image to obtain the final pantograph slide image.

[0025] To avoid the problem of the complete pantograph sling being segmented by two adjacent images when acquiring the final pantograph sling image, this invention performs a pantograph sling integrity determination. If the integrity is not satisfied, it indicates that the complete pantograph sling is segmented by two adjacent images. In this case, considering the difference between whether the radar detected the passing vehicle in advance and whether subsequent image stretching and compression are necessary, this invention directly stitches together the images containing the two parts of the pantograph sling before segmenting them, displaying the pantograph sling completely on a single image when the radar did not detect the passing vehicle in advance (when subsequent image stretching and compression are not necessary). For cases where the radar did not detect the passing vehicle in advance (when subsequent image stretching and compression are not necessary), this invention stitches together five images, performs stretching and compression operations, and then segments them, regardless of whether the complete pantograph sling is segmented by two adjacent images. This ensures that the pantograph sling is displayed completely on a single image, and the original proportions of the pantograph sling in the image are consistent with the original image, with no stretching or compression issues. Furthermore, when the stitched image needs to be segmented, this invention segments the stitched image using the geometric center of the pantograph slider as the image center, obtaining the final pantograph slider image. This ensures that the segmented pantograph slider is centered in the image and that the image width remains consistent with the original image.

[0026] Furthermore, in step 1), when the photoelectric sensor module detects an approaching vehicle, it simultaneously controls the vehicle number capturing module to take a picture; the vehicle number capturing module is a color camera and uses a fixed frame rate to capture vehicle number images.

[0027] The vehicle number shooting module in the method of this invention uses a high frame rate color camera instead of a black and white camera to increase the color contrast between the vehicle background and the vehicle number, thus solving the problem of the vehicle number not being distinguishable when the foreground and background are blended.

[0028] Furthermore, in step 1), the time for the train number capturing module to take pictures is controlled according to the maximum length of the first car and the radar measured speed, so that only the image of the first car is collected.

[0029] When the train type is a high-speed train, the pantograph imaging module takes pictures based on the maximum length of a single train and the radar measured speed, so that the images of the high-speed train are only collected when the high-speed train passes by.

[0030] When the train type is a conventional speed train, the pantograph imaging module's photo-taking time is controlled according to the maximum length of the first car and the radar's measured speed, so that only the image of the first car is collected.

[0031] This invention takes into account that the train numbers of both conventional and high-speed trains are located on the lead car. By controlling the acquisition time of the train number camera, images of only the lead car are captured, significantly reducing the amount of image data and improving the subsequent recognition and processing speed. Furthermore, this invention also considers that the pantograph of conventional trains is located on the lead locomotive. By controlling the camera's acquisition time, images of only the lead car are captured, significantly reducing the amount of image data and improving the subsequent recognition and processing speed. For high-speed trains, a full-vehicle shooting method is used, controlling the acquisition time of the high-speed pantograph acquisition module to obtain images of all pantographs of the entire high-speed train. Therefore, this invention ensures image completeness while reducing the number of images acquired, thus avoiding the problems of excessive bandwidth consumption and long transmission times in densely trafficked sections, which could affect the system's subsequent train reception.

[0032] To achieve the above objectives, the present invention also provides a pantograph sliding image acquisition system, including a photoelectric sensor for monitoring oncoming vehicles and a radar, wherein the photoelectric sensor is oriented towards the oncoming vehicle, and further including a pantograph imaging module for acquiring pantograph sliding images and a vehicle number imaging module for acquiring vehicle number images, and further including a memory and a processor, and computer program instructions stored in the memory and running on the processor, wherein the processor is used to execute the computer program instructions stored in the memory to implement the pantograph sliding image acquisition method steps as described above, and achieve the same beneficial effects as the method.

[0033] Furthermore, the photoelectric sensor is mounted on the railside column at a 30-degree angle to the direction of travel.

[0034] The system of this invention is set at this angle to ensure that the information of the approaching train is detected before the train arrives at the equipment, and the photoelectric sensor module starts the equipment to start working immediately after detecting the approaching train.

[0035] Furthermore, the pantograph imaging module is installed on the top of the single column or on a rigid cross span. If it is installed on the top of the column, it takes a downward view at a fixed angle to the column, with a vertical distance of 2.5m from the pantograph. If it is installed on the rigid cross span, it takes a vertical downward view, with a vertical distance of 2.5m from the pantograph.

[0036] The system of this invention features a design for the pantograph imaging module that allows the lens to completely cover the entire pantograph (including the double sliding plates, pantograph angles, and insulators) at this distance, while ensuring sufficiently high image resolution to clearly identify millimeter-level wear or foreign objects. Furthermore, considering that the pantograph will rise and fall with the changing height of the overhead contact line during train operation, the equipment installation must allow for sufficient safety margin beyond the maximum pantograph lifting height. Therefore, a vertical distance of 2.5 meters ensures that even when the pantograph is raised to its highest point or experiences severe vibrations, the camera and its protective cover will not physically collide with the pantograph, guaranteeing the safety of both the equipment and the train.

[0037] The above description is merely an overview of the technical solution of the present invention. In order to better understand the technical means of the present invention and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of the present invention more apparent and understandable, preferred embodiments are described in detail below with reference to the accompanying drawings. Attached Figure Description

[0038] Figure 1 This is an installation diagram of the system in this embodiment when applied to a single-column configuration;

[0039] Figure 2 This is an installation diagram of the system in this embodiment when applied to a rigid crossover configuration;

[0040] Figure 3 This is a schematic diagram of the installation of the photoelectric sensor module used in this embodiment. Detailed Implementation

[0041] The technical solution of the present invention will be clearly and completely described below with reference to specific embodiments. However, those skilled in the art should understand that the embodiments described below are only for illustrating the present invention and should not be regarded as limiting the scope of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0042] Example of pantograph sliding plate image acquisition system

[0043] The innovation of this embodiment lies in:

[0044] 1. The existing system uses radar to detect passing trains. For high-speed trains exceeding 350 km / h, the system may miss vehicle information due to untimely detection. Introducing photoelectric sensors, forming a dual-switch detection mechanism with radar, ensures reliable and early detection of approaching trains even at high speeds, thus resolving the problem of missed vehicle information detection.

[0045] 2. An image adaptive processing algorithm is introduced. When a high-speed train passes by, the photoelectric sensor detects the train's speed. The algorithm combines the system's preset speed with the radar's final measured speed to dynamically calculate the image stretching and compression ratio. The original image is then corrected to ensure that the final image is displayed normally and that key information such as the pantograph's sliding plate is centered in the image. This solves the problem that high speeds can cause image stretching or compression distortion, and the pantograph's sliding plate may not be centered in the image.

[0046] 3. By employing intelligent image recognition algorithms and combining them with business logic, the images of train numbers and pantograph contact plates are intelligently filtered, reducing the amount of image data and compressing the final uploaded valid image data to a maximum of 5 images (maximum 3 images of train numbers + 2 images of pantograph contact plates). This greatly improves transmission efficiency and reduces the impact of current train passage on subsequent train passage processing on lines with frequent train passages.

[0047] 4. For vehicles with blue bodies and red license plates, monochrome cameras cannot effectively distinguish the license plate from the background, resulting in the license plate "blending" and being difficult to identify. Replacing the monochrome camera with a color camera utilizes color information to clearly separate the red license plate from the blue vehicle body background, fundamentally solving the difficulty of license plate recognition for this type of vehicle.

[0048] 5. Based on the key images (simplified vehicle number and pantograph contactor image) selected by the recognition algorithm, the platform's display interface is integrated and simplified. The vehicle number and key pantograph contactor image information are directly and centrally displayed on a single interface. This allows inspectors to quickly obtain core information, significantly reducing workload and improving inspection efficiency.

[0049] This embodiment of the system includes a photoelectric sensor and a radar for monitoring oncoming vehicles. The photoelectric sensor is oriented towards the direction of the oncoming vehicle. It also includes a pantograph imaging module for acquiring images of the pantograph slider and a vehicle number imaging module for acquiring images of vehicle license plates. It also includes a memory and a processor, as well as computer program instructions stored in the memory and running on the processor. The processor is used to execute the computer program instructions stored in the memory to implement the pantograph slider image acquisition method steps.

[0050] Based on the site environment, the pantograph imaging module in this embodiment is installed in two ways: single-column and rigid cross-span. For example... Figure 1As shown, in the single-column configuration: the pantograph imaging module is installed at the top of the column, forming a fixed angle with the column to take downward-facing shots. The plane of the module is perpendicular to the direction of travel, and its vertical distance from the pantograph is approximately 2.5 meters. Figure 2 As shown, in the rigid cross-span configuration: the pantograph imaging module is installed directly above the center of the track, fixed on the rigid cross-span, and shoots vertically downwards, approximately 2.5 meters away from the pantograph. Regarding the setting of the photoelectric sensor to ensure that it detects incoming train information before the train reaches the equipment, in this embodiment, the photoelectric sensor module is installed on the railside pole at a 30-degree angle to the direction of travel, ensuring that it detects incoming train information before the train reaches the equipment.

[0051] In this embodiment, for vehicles with blue bodies and red license plates, the system addresses the issue of vehicle number and background blending by using a high frame rate color camera instead of a monochrome camera. This increases the color contrast between the vehicle body background and the license plate, resolving the problem of the license plate not being distinguishable when the foreground and background are blended.

[0052] The following describes the steps of the pantograph sliding plate image acquisition method implemented in this embodiment:

[0053] To avoid the problem of high-speed train radar missing vehicles, this embodiment adds a photoelectric sensor module. The photoelectric sensor module is mounted on a railside pole at a 30-degree angle to the direction of travel, ensuring that it detects approaching vehicles before the train reaches the equipment. Once the photoelectric sensor module detects an approaching vehicle, it immediately starts the equipment to begin operation. Specifically:

[0054] 1. If the radar also detects the vehicle in advance, use radar speed measurement to control the camera to take pictures;

[0055] 2. If the radar fails to detect the passing train in advance, resulting in a missed detection of the train ahead, since the passing train type and speed are basically the same at fixed locations and times, the camera with the current speed control configured by the system takes a picture. After the passing train is over, the stretching and compression ratio of the image is calculated based on the final speed detected by the radar and the speed configured by the system. The pantograph sliding plate image after recognition is then stretched, compressed and segmented. Assuming the radar's final detection speed is v1 and the system's current time-period configuration speed is v2, the image stretching / compression ratio is calculated as r = v1 / v2. The image number of the pantograph slider after identification is i, and the width of a single original image is w. The five images (i-2, i-1, i, i+1, i+2) are then stitched together into a single large image with a width of w*5. After stretching / compression, the image width is w1 = r*w*5. Using the geometric center of the stretched / compressed pantograph slider as the origin, the stretched / compressed image is segmented. A width of w / 2 pixels is taken before and after the origin coordinates to segment the pantograph slider. After segmentation, the pantograph slider is centered in the image, and the image width remains consistent with the original image. Furthermore, the pantograph slider's original proportions in the image are consistent with the original image, and there is no stretching / compression issue in the image display.

[0056] This embodiment addresses the issue of large image data volume and long transmission time of vehicle license plate numbers affecting subsequent vehicle passages. It employs a high frame rate camera combined with a modulated light source for supplemental lighting. During vehicle passage, a fixed frame rate of 100 frames per second is used to capture clear images of the vehicle license plate number, ensuring that a single image capture can cover the entire license plate number. Specifically:

[0057] Both conventional and high-speed trains have their car numbers located on the lead car. By controlling the acquisition time of the car number camera, images of only the lead car are captured, significantly reducing the amount of image data and improving the speed of subsequent recognition and processing. The maximum length of the lead car is taken as L meters, the radar measurement speed is V meters per second, and the acquisition time is T = L / V. The camera automatically stops acquiring images after time T is reached.

[0058] The acquired vehicle license plate images are identified, and three images with complete license plate numbers are selected according to the corresponding algorithm, further reducing the number of license plate images to three. The main processing steps are as follows: a. Image enhancement: A histogram local equalization algorithm is used to stretch the grayscale range of the image, increase the contrast between the license plate number and the background, and improve the brightness of darker areas; b. License plate region localization: The license plate region in the image is located. The license plate number is generally painted in a fixed position on the vehicle body. After the camera angle is fixed, the position of the license plate number in the image is basically consistent. We define a rough approximate area to coarsely locate the license plate number, narrowing the processing range and improving processing efficiency. Based on accuracy and complex backgrounds, the YOLOv5 model, with its higher efficiency and robustness, is used to quickly locate the precise position of the train number and derive the coordinates of its four corner points. c. Train Number Region Correction: The detected train number region image is cropped, and the position is corrected using a perspective algorithm based on the coordinates of the four corner points. d. Train Number Recognition: A CRNN network is used to recognize the train number and derive the character content of the train number region. e. Train Number Verification and Image Selection: The recognized train numbers are verified according to the train number encoding rules (e.g., character type, number of characters, alphanumeric position information, etc.). The train number image that meets the encoding rules and has the highest confidence is selected. Once the number of selected images reaches a maximum of 3, the recognition of subsequent train number images is stopped. This reduces thousands of train number images to a maximum of 3 images, reducing data volume, improving transmission efficiency, and mitigating the impact on subsequent trains on lines with frequent train passages.

[0059] For pantograph imaging, this embodiment employs the following process to ensure the final image is displayed correctly and the pantograph slider is fully visible in a single image:

[0060] A high-speed pantograph acquisition module supporting real-time data compression is employed to acquire and store images at high speed and real-time at a frame rate matching the train's speed after detecting an approaching vehicle. For high-speed trains, a full-vehicle imaging method is used, controlling the acquisition time of the pantograph's high-speed acquisition module to capture images of all 16 train sets. The maximum length of a single car is defined as L1 meters, the radar measurement speed as V meters per second, and the acquisition time T1 = L1 / V. The camera automatically stops acquiring images after time T1. For conventional trains, the pantograph is located on the lead locomotive. By controlling the camera's acquisition time, only images of the lead car are captured, significantly reducing the amount of image data and improving the subsequent recognition and processing speed. The maximum length of the lead car is defined as L2 meters, the radar measurement speed as V meters per second, and the acquisition time T2 = L2 / V. The camera automatically stops acquiring images after time T2.

[0061] Pantograph Sliding Plate Recognition Scheme: The scheme involves recognizing the acquired pantograph sliding plate images, selecting images with the complete pantograph sliding plate based on a corresponding algorithm, and storing and transmitting these images. The main processing steps are as follows: a. Image Enhancement: A histogram local equalization algorithm is used to stretch the image's grayscale range, increase image contrast, and improve the brightness of darker areas; b. Pantograph Positioning: Once the installation scheme is determined, the pantograph sliding plate positions in the vertical direction of the image are basically consistent. Therefore, a fixed area in the middle of the image can be set as the recognition area to reduce interference from non-information areas on both sides of the image. The pantograph slider region of a single image is detected using the YOLOv5 model, and the coordinate information of the pantograph slider is obtained; c. Pantograph slider integrity determination: The height of the pantograph slider frame is calculated. If the height of the detected pantograph slider is less than the set threshold, the pantograph slider is determined to be incomplete in the image; d. Secondary detection of pantograph slider: (1) Radar detects passing vehicles in advance: If the pantograph slider is incomplete in the original image, the previous or next image of the current image is stitched together based on whether the pantograph slider appears at the left head position or the right tail position of the image. The YOLOv5 model is used again to detect and obtain the complete pantograph slider coordinate information. Based on the coordinate information, the geometric center position of the pantograph slider is calculated. The image is segmented based on the resolution size of the original image. The size of the segmented image is the same as that of the original image, and the pantograph slider is located at the center of the image. (2) The radar did not detect the vehicle in advance: Based on the current pantograph sliding plate image, take 2 images in front and 2 images behind, and stitch together a total of 5 images. Perform stretching and compression operation on the stitched image according to the calculated stretching and compression ratio. Take the geometric center of the pantograph sliding plate in the image after stretching and compression as the origin, and perform segmentation operation on the stretched and compressed image. The size of the segmented image is almost the same as the original image. The pantograph sliding plate is located in the center of the image. At the same time, the display ratio of the pantograph sliding plate in the image is consistent with the original image. There is no stretching and compression problem; e. The storage and transmission only includes the image of the pantograph sliding plate as the effective data, which greatly reduces the amount of data and improves the transmission efficiency.

[0062] This embodiment achieves the following effects:

[0063] 1. Introduce photoelectric sensors to form a dual-switch vehicle detection mechanism with radar to detect vehicles that are missed by radar on high-speed trains;

[0064] 2. Using photoelectric sensor signals as triggers, combined with the system's preset speed and the radar's measured speed, the image stretching / compression ratio is dynamically calculated to obtain a complete and normal pantograph sliding plate image;

[0065] 3. Use an image adaptive processing algorithm to correct the original image to ensure that the final image is displayed normally and that the key pantograph sliding plate information is always located in the center of the image;

[0066] 4. By applying intelligent recognition algorithms, the number of images is greatly reduced, transmission efficiency is improved, and interference with subsequent vehicle processing is reduced;

[0067] 5. Using a color industrial camera, the red license plate number can be clearly separated from the blue vehicle body background, solving the difficulty of identifying and distinguishing license plate numbers for this type of vehicle.

[0068] Example of pantograph sliding plate image acquisition method

[0069] The method in this embodiment is based on the pantograph and skateboard image acquisition system described above to implement the pantograph and skateboard image acquisition method steps. The specific implementation and the effects achieved are described in detail in the pantograph and skateboard image acquisition system embodiment above, and will not be repeated here.

[0070] Although the present invention has been described in detail above with general descriptions and specific embodiments, modifications or improvements can be made to it, which will be obvious to those skilled in the art. Therefore, all such modifications or improvements made without departing from the spirit of the present invention fall within the scope of protection claimed by the present invention.

Claims

1. A method for acquiring images of a pantograph sliding plate, characterized in that, The method introduces a photoelectric sensor to form a dual-switch vehicle detection mechanism with radar. The photoelectric sensor is oriented towards the oncoming vehicle. The method includes the following steps: 1) Triggered by the photoelectric sensor signal, when the photoelectric sensor module detects an approaching vehicle, it controls the pantograph imaging module to take pictures based on the system's preset speed and the radar's measured speed. 2) After the vehicle passes, the image stretching and compression ratio is calculated based on the final speed detected by the radar and the preset speed of the system. The identified pantograph sliding plate image is then stretched, compressed, and segmented to obtain the final pantograph sliding plate image.

2. The pantograph sliding plate image acquisition method according to claim 1, characterized in that, In step 1), the pantograph imaging module is controlled to take pictures by combining the system's preset speed and the radar's measured speed, including: When the photoelectric sensor module detects an approaching vehicle, if the radar has also detected the vehicle in advance, the radar speed measurement will be used to control the pantograph camera module to take a picture. When the photoelectric sensor module detects an approaching vehicle, if the radar has not detected the vehicle beforehand, the pantograph camera module will take a picture using the system's preset speed control.

3. The pantograph sliding plate image acquisition method according to claim 2, characterized in that, In step 1), the system preset speed is a speed configured for different time periods.

4. The pantograph sliding plate image acquisition method according to claim 2, characterized in that, In step 2), the calculation of the image stretching and compression ratio based on the final speed detected by the radar and the preset speed of the system includes: if the radar also detected the passing vehicle in advance, the stretching and compression ratio is 1; if the radar did not detect the process in advance, the stretching and compression ratio is the ratio of the radar's final detected speed to the preset speed of the system.

5. The pantograph sliding plate image acquisition method according to claim 2, characterized in that, In step 2), the pantograph sliding plate image after recognition is stretched, compressed and segmented to obtain the final pantograph sliding plate image, including: image enhancement of the image obtained by the pantograph imaging module, detection of a single image containing the pantograph sliding plate area, and determination of the pantograph sliding plate integrity of the single image. If the completeness meets the threshold and the radar has also detected the passing vehicle in advance, then the single image is determined as the final pantograph skateboard image. If the integrity does not meet the threshold, and the radar has detected the passing vehicle in advance, then the previous or next image of the single image is stitched together depending on whether the pantograph sliding plate area is located at the left head position or the right tail position of the single image. The pantograph sliding plate area is located in the stitched image to obtain the geometric center position of the pantograph sliding plate. Based on the resolution size of the single image, the stitched image is segmented with the geometric center position as the image center position to obtain the final pantograph sliding plate image. If the radar does not detect the vehicle in advance, then based on the single image, take two images in front and two images behind, for a total of five images, and stitch them together. Then, perform a stretching and compression operation on the stitched image according to the stretching and compression ratio. Using the geometric center of the stretched and compressed pantograph slide in the image as the origin and the resolution size of the single image as the reference, perform a segmentation operation on the stretched and compressed image to obtain the final pantograph slide image.

6. The pantograph sliding plate image acquisition method according to claim 1, characterized in that, In step 1), when the photoelectric sensor module detects an approaching vehicle, it simultaneously controls the license plate number capturing module to take a picture; the license plate number capturing module is a color camera and uses a fixed frame rate to capture license plate number images.

7. The pantograph sliding plate image acquisition method according to claim 6, characterized in that, In step 1), the time for the car number capturing module to take pictures is controlled according to the maximum length of the first car and the radar measured speed, so that only the image of the first car is collected. When the train type is a high-speed train, the pantograph imaging module takes pictures based on the maximum length of a single train and the radar measured speed, so that the images of the high-speed train are only collected when the high-speed train passes by. When the train type is a conventional speed train, the pantograph imaging module's photo-taking time is controlled according to the maximum length of the first car and the radar's measured speed, so that only the image of the first car is collected.

8. A pantograph sliding plate image acquisition system, characterized in that, The method includes a photoelectric sensor for monitoring oncoming vehicles and a radar, the photoelectric sensor being oriented towards the oncoming vehicle, a pantograph imaging module for acquiring images of the pantograph slider and a vehicle number imaging module for acquiring images of vehicle license plates, a memory and a processor, and computer program instructions stored in the memory and running on the processor, the processor being used to execute the computer program instructions stored in the memory to implement the pantograph slider image acquisition method steps as described in any one of claims 1 to 7.

9. The pantograph sliding plate image acquisition system according to claim 8, characterized in that, The photoelectric sensor is installed on the railside column at a 30-degree angle to the direction of travel.

10. The pantograph sliding plate image acquisition system according to claim 8, characterized in that, The pantograph imaging module is installed on the top of the single column or on a rigid cross span. If it is installed on the top of the column, it takes a downward view at a fixed angle to the column, with a vertical distance of 2.5m from the pantograph. If it is installed on the rigid cross span, it takes a downward view vertically, with a vertical distance of 2.5m from the pantograph.

Images