A vehicle-mounted non-contact contact wire suspension state detection method and system

Abstract

The present invention discloses a vehicle-mounted non-contact contact line suspension state detection method and system, the method is applied to a first array camera and a second array camera arranged in the direction of the contact line suspension device, including the steps of: controlling the first array camera to take a first picture at a preset time interval, and sending the first picture to the vehicle-mounted industrial control computer in real time; identifying whether there is a key feature of the suspension device in the first picture by the vehicle-mounted industrial control computer; if so, controlling the second array camera by the vehicle-mounted industrial control computer to take a second picture of the suspension state of the contact line, and controlling the second array camera to send the second picture to the vehicle-mounted industrial control computer in real time; detecting whether there is an abnormality in the suspension state of the contact line in the second picture by the vehicle-mounted industrial control computer. The present invention realizes accurate detection of the vehicle-mounted non-contact contact line suspension device during train operation, which is convenient for timely discovery of the failure of the suspension device and ensures the safe operation of the train.

Description

Technical Field

[0001] The present invention relates to the technical field of railway contact network detection, and in particular to a vehicle-mounted non-contact contact wire suspension state detection method and system. Background Art

[0002] In recent years, the rail transit industry has developed rapidly. Subways, light rails and high-speed rails have become an indispensable means of transportation for people to go out every day. In the construction of such rail transit modes, how to ensure the long-term safe operation of trains is an extremely important maintenance plan. Among them, the train suspension device has many parts and types, and it plays the role of fixing the contact line. Long-term vibration during train operation can easily cause it to malfunction. In addition, since the suspension device is generally located at a high position, it is difficult for staff to observe it clearly from the ground, which increases the difficulty of manual inspection.

[0003] One of the current detection methods is to install a camera on the roof of the train to collect the suspension status of the contact line around the clock and save the collected video data in the database. However, this detection method requires a large amount of storage space to be reserved in the database for storing the video data of the suspension status of the contact line, and it is difficult to accurately locate the position of the faulty suspension device when analyzing the suspension status of the contact line. Another method is to set up an infrared laser trigger device to take a photo when the trigger device detects the suspension device, but the accuracy of the infrared laser trigger device is greatly affected by environmental factors and cannot guarantee the accuracy of the contact line suspension status detection.

[0004] Therefore, there is a need for a method for detecting the status of a vehicle-mounted non-contact contact wire suspension, which can realize accurate detection of the vehicle-mounted non-contact contact wire suspension device during train operation, facilitate analysis of the fault conditions of the suspension device, and enable maintenance personnel to respond quickly and perform maintenance in a timely manner when a suspension device fails, thereby ensuring safe operation of the train. Summary of the invention

[0005] In order to solve the technical problem that it is difficult to accurately detect the on-board non-contact contact wire suspension device during train operation, the present invention provides a method and system for detecting the suspension state of a on-board non-contact contact wire. The specific technical solution is as follows:

[0006] The present invention provides a vehicle-mounted non-contact contact line suspension state detection method, which is applied to a first area array camera and a second area array camera arranged in the direction of a suspension device of the contact line, comprising the steps of:

[0007] Controlling the first area array camera to capture a first picture at a preset time interval, and sending the first picture to the vehicle-mounted industrial computer in real time;

[0008] identifying, by the on-board industrial computer, whether the key features of the suspension device exist in the first image;

[0009] If yes, the vehicle-mounted industrial computer controls the second area array camera to capture the suspension state of the contact line to obtain a second picture, and controls the second area array camera to send the second picture to the vehicle-mounted industrial computer in real time;

[0010] The vehicle-mounted industrial computer is used to detect whether there is any abnormality in the suspension state of the contact line in the second picture.

[0011] The vehicle-mounted non-contact contact wire suspension state detection method provided by the present invention sets two kinds of area array cameras. When the industrial control computer determines that the first area array camera has photographed the suspension device, it controls the second area array camera to photograph the current suspension state of the contact wire, and specifically identifies whether the parts are intact based on the picture taken by the second area array camera. According to the shooting result of the first area array camera, it can be identified whether there is a suspension device at the current position, and there is no need to reserve a large amount of storage space to store the video data of the suspension state of the contact wire photographed by the first area array camera, thereby improving the accuracy and convenience of detecting the suspension state of the vehicle-mounted non-contact contact wire during train travel.

[0012] In some embodiments, the identifying, by the onboard industrial computer, whether the key feature of the suspension device exists in the first image specifically includes:

[0013] Identify, by the vehicle-mounted industrial computer, whether there is a first straight line perpendicular to the x-axis of the image coordinate system, a second straight line horizontal to the x-axis of the image coordinate system, and a third straight line at an oblique angle to the x-axis of the image coordinate system in the first image;

[0014] If the judgment result is yes, secondly judge whether the first image includes a first sub-feature and a second sub-feature, the first sub-feature being that the lengths of the first straight line, the second straight line and the third straight line are all within the corresponding preset length threshold, and the second sub-feature being that there is an intersection between any two of the first straight line, the second straight line and the third straight line;

[0015] If the secondary determination result is yes, it is determined that the key features of the suspension device exist in the first picture.

[0016] The vehicle-mounted non-contact contact line suspension state detection method provided by the present invention discloses a method for identifying straight line features in a first image to determine whether there are key features of a suspension device in the first image, balance the accuracy and efficiency of identifying the suspension device in the first image, and facilitate real-time control of a second array camera to collect images.

[0017] In some embodiments, the identifying, by the onboard industrial computer, whether the key features of the suspension device exist in the first image specifically includes:

[0018] The vehicle-mounted industrial computer compares the first image with a plurality of image templates of the contact wire suspension device pre-stored in the vehicle-mounted industrial computer, and calculates the image matching degree between the first image and each of the image templates;

[0019] If the picture matching degree is greater than a preset first matching degree threshold, it is determined that the key feature of the suspension device exists in the first picture.

[0020] The vehicle-mounted non-contact contact wire suspension state detection method provided by the present invention discloses another solution for judging whether key features of a suspension device exist in a first image by comparing image templates.

[0021] In some embodiments, the detecting, by the on-board industrial computer, whether there is an abnormality in the suspension state of the contact line in the second picture, specifically includes:

[0022] Determine whether there is a fourth straight line perpendicular to the x-axis of the image coordinate system, a fifth straight line horizontal to the x-axis of the image coordinate system, and a sixth straight line at an oblique angle to the x-axis of the image coordinate system in the second picture;

[0023] If the judgment result is yes, secondly judge whether the second picture includes the third sub-feature, the fourth sub-feature and the fifth sub-feature, the third sub-feature is that the lengths of the fourth straight line, the fifth straight line and the sixth straight line are all within the corresponding preset length threshold, the fourth sub-feature is that the slopes of the fourth straight line, the fifth straight line and the sixth straight line are all within the corresponding preset slope threshold, and the fifth sub-feature is that there is an intersection between any two of the fourth straight line, the fifth straight line and the sixth straight line;

[0024] If the secondary judgment result is yes, taking the intersection point between any two straight lines among the fourth straight line, the fifth straight line and the sixth straight line as a key point, and taking each of the key points as the center in the second picture, intercepting an area of ​​a preset pixel area to generate a key point picture;

[0025] Comparing each of the key point images with the parts data images stored in the database of the vehicle-mounted industrial computer to obtain a first comparison result;

[0026] Calculating the part distance between each of the key points in the second image, and after standardizing the part distance, comparing the part distance with the distance data stored in the database of the vehicle-mounted industrial computer to obtain a second comparison result;

[0027] It is determined whether the suspension state of the contact line is abnormal according to the first comparison result and the second comparison result.

[0028] The vehicle-mounted non-contact contact line suspension state detection method provided by the present invention discloses a scheme for identifying whether there is an abnormality in the suspension state of the contact line in a second image. By comparing the similarity between the key point image and the part data image in the second image, as well as the part distance between the key points and the distance data stored in the database, it is judged whether the suspension state of the contact line is abnormal. The suspension state of the contact line is judged comprehensively based on multiple factors to improve the accuracy of detection.

[0029] In some embodiments, after the region of the preset pixel area is intercepted to generate the key point image, before the first comparison result is obtained by comparing each of the key point images with the part image data stored in the database of the vehicle-mounted industrial computer, the method further includes:

[0030] Performing edge recognition on the key point image;

[0031] The key point images are screened according to screening features, where the screening features include the area of ​​the key point images, the aspect ratio of the key point images, and the slope of the edge of the key point images.

[0032] The vehicle-mounted non-contact contact wire suspension state detection method provided by the present invention screens key point images through edge recognition, thereby further improving the accuracy of contact wire suspension state detection.

[0033] In some embodiments, after judging whether the suspension state of the contact line is abnormal according to the first comparison result and the second comparison result, the method further includes:

[0034] If the judgment result is yes, again compare each of the key point images with the part image data stored in the database of the vehicle-mounted industrial computer to obtain a third comparison result, and judge whether each of the third comparison results is greater than a preset similarity threshold;

[0035] If there is the third comparison result that is not greater than the similarity threshold, generating a parts missing alarm message;

[0036] If each of the third comparison results is greater than the similarity threshold, the second image is compared with the suspension state image stored in the database of the vehicle-mounted industrial computer to obtain a second matching degree, and the accuracy of the abnormal detection result of the suspension state of the contact line is verified based on the second matching degree.

[0037] The vehicle-mounted non-contact contact wire suspension state detection method provided by the present invention compares the key point image with the part image again after determining that the contact wire suspension state is abnormal, so as to avoid calculation errors in the first comparison process. When the similarity between the key point image and the part image is greater than the similarity threshold, the second image is compared with the suspension state image stored in the database. Multiple verifications greatly improve the accuracy of the abnormality detection results.

[0038] In some implementations, verifying the accuracy of the abnormality detection result of the suspension state of the contact line according to the second matching degree specifically includes:

[0039] If the second matching degree is greater than a preset second matching degree threshold, the abnormality detection result of the suspension state of the contact line is updated, it is determined that there is no abnormality in the suspension state of the contact line, and the second image, the key point image corresponding to the second image, and the part distance corresponding to the second image are stored in the database of the on-board industrial computer;

[0040] If the second matching degree is not greater than the second matching degree threshold, the part missing alarm information is generated.

[0041] The vehicle-mounted non-contact contact wire suspension state detection method provided by the present invention discloses a scheme for verifying the accuracy of the abnormal detection result of the suspension state of the contact wire according to a second matching degree obtained by comparing a second image with a suspension state image stored in a database of a vehicle-mounted industrial computer, and continuously improving the accuracy and detection speed of the vehicle-mounted non-contact contact wire suspension state detection by expanding the database and training the recognition model.

[0042] In some embodiments, a constantly on LED is provided to cooperate with the first area array camera, and a flashing LED is provided to cooperate with the second area array camera;

[0043] When the on-board industrial computer controls the second area array camera to capture the second picture, the flashing LED is controlled to be in working state at the same time.

[0044] The vehicle-mounted non-contact contact wire suspension state detection method provided by the present invention prevents the flashing LED from always being in a working state and causing an impact on the safety of train operation.

[0045] In some implementations, the vehicle-mounted industrial computer receives acceleration data sent by an acceleration sensor in real time;

[0046] After the on-board industrial computer detects whether the suspension state of the contact line in the second picture is abnormal, the method further includes:

[0047] When an abnormality is detected, the onboard industrial computer is controlled to calculate the current position of the train according to the acceleration data, and the number of the suspension device of the contact line where the abnormality occurs is determined according to the current position of the train.

[0048] The vehicle-mounted non-contact contact wire suspension state detection method provided by the present invention receives acceleration data sent by an acceleration sensor in real time. After detecting that the suspension state is abnormal, the current position of the train is determined according to the acceleration data, and an early warning is issued in time according to the location and number of the suspension device fault, so as to facilitate timely maintenance of the suspension device.

[0049] In some embodiments, the present invention further provides a vehicle-mounted non-contact contact wire suspension state detection system, comprising:

[0050] A first area array camera, which is arranged toward the suspension device of the contact line and is used to take a first picture at a preset time interval and send the first picture to the vehicle-mounted industrial computer in real time;

[0051] The vehicle-mounted industrial computer is connected to the first area array camera and is used to identify whether the key features of the suspension device exist in the first image;

[0052] A second area array camera is connected to the vehicle-mounted industrial computer, and is arranged toward the suspension device of the contact wire. The second area array camera is used to capture the suspension state of the contact wire to obtain a second image under the control of the vehicle-mounted industrial computer when the vehicle-mounted industrial computer recognizes that there are key features of the suspension device in the first image, and send the second image to the vehicle-mounted industrial computer in real time.

[0053] The vehicle-mounted industrial computer is further configured to detect whether there is any abnormality in the suspension state of the contact line in the second image after receiving the second image.

[0054] The present invention provides a vehicle-mounted non-contact contact wire suspension state detection method and system, which includes at least one of the following technical effects:

[0055] (1) By setting two area array cameras, when the industrial computer determines that the first area array camera has photographed the suspension device, the second area array camera is controlled to photograph the current suspension state of the contact line, and whether the parts are intact is specifically identified based on the image photographed by the second area array camera. According to the photographing result of the first area array camera, it is possible to identify whether there is a suspension device at the current position, and there is no need to reserve a large amount of storage space to store the video data of the suspension state of the contact line photographed by the first area array camera, thereby improving the accuracy and convenience of detecting the suspension state of the on-board non-contact contact line during the running of the train;

[0056] (2) by identifying the straight line features in the first image, determining whether there are key features of the suspension device in the first image, balancing the accuracy and efficiency of identifying the suspension device in the first image, and facilitating real-time control of the second area array camera to collect images;

[0057] (3) By comparing the similarity between the key point image and the part data image in the second image, as well as the part distance between the key points and the distance data stored in the database, it is determined whether the hanging state of the contact line is abnormal, and the hanging state of the contact line is determined comprehensively based on multiple factors to improve the accuracy of detection;

[0058] (4) Screen key point images through edge recognition to further improve the accuracy of contact line suspension state detection;

[0059] (5) After determining that the contact line suspension state is abnormal, the key point image is compared with the part image again to avoid calculation errors in the first comparison process. When the similarity between the key point image and the part image is greater than the similarity threshold, the second image is compared with the suspension state image stored in the database. Multiple verifications greatly improve the accuracy of the abnormality inspection results;

[0060] (6) By comparing the second image with the suspension state image stored in the database of the on-board industrial computer to obtain a second matching degree, the accuracy of the abnormal detection result of the suspension state of the contact line is verified, and the accuracy and detection speed of the on-board non-contact contact line suspension state detection are continuously improved according to the expansion of the database and the training of the recognition model;

[0061] (7) By receiving the acceleration data sent by the acceleration sensor in real time, after detecting an abnormal suspension state, the current position of the train is determined based on the acceleration data, and an early warning is issued in a timely manner based on the location and number of the suspension device fault, so as to facilitate timely maintenance of the suspension device. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0063] Figure 1 This is a flow chart of a vehicle-mounted non-contact contact wire suspension state detection method of the present invention;

[0064] Figure 2 A flow chart for identifying key features of a suspension device in a first picture in a vehicle-mounted non-contact contact wire suspension state detection method of the present invention;

[0065] Figure 3 Another flow chart for identifying key features of the suspension device in the first picture in a vehicle-mounted non-contact contact wire suspension state detection method of the present invention;

[0066] Figure 4 A flow chart of identifying whether the suspension state in the second picture is abnormal in a vehicle-mounted non-contact contact wire suspension state detection method of the present invention;

[0067] Figure 5 A flow chart for identifying whether the suspension state in the second picture is abnormal in a vehicle-mounted non-contact contact wire suspension state detection method of the present invention;

[0068] Figure 6 Another flow chart of identifying whether the suspension state in the second picture is abnormal in a vehicle-mounted non-contact contact wire suspension state detection method of the present invention;

[0069] Figure 7 This is an example diagram of the hardware structure required for a vehicle-mounted non-contact contact wire suspension state detection method of the present invention;

[0070] Figure 8 Another flow chart of a vehicle-mounted non-contact contact wire suspension state detection method of the present invention;

[0071] Fig. 9 This is an example diagram of a vehicle-mounted non-contact contact wire suspension state detection system of the present invention.

[0072] Numbers in the figure: first area array camera - 1, second area array camera - 2, constant-on LED - 3, flashing LED - 4, acceleration sensor - 5, first area array camera - 10, vehicle-mounted industrial computer - 20 and second area array camera - 30. DETAILED DESCRIPTION

[0073] In the following description, specific details such as specific system structures, technologies, etc. are provided for the purpose of illustration rather than limitation, so as to provide a thorough understanding of the embodiments of the present application. However, it should be clear to those skilled in the art that the present application may also be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to prevent unnecessary details from obstructing the description of the present application.

[0074] It should be understood that when used in this specification and the appended claims, the term "comprising" indicates the presence of the described features, integers, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or collections.

[0075] In order to simplify the drawings, only the parts related to the present invention are schematically shown in each figure, and they do not represent the actual structure of the product. In addition, in order to simplify the drawings and facilitate understanding, in some figures, only one of the parts with the same structure or function is schematically drawn or marked. In this article, "one" not only means "only one", but also means "more than one".

[0076] It should be further understood that the term “and / or” used in the specification and appended claims refers to any combination and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0077] In addition, in the description of the present application, the terms "first", "second", etc. are only used to distinguish the description and cannot be understood as indicating or implying relative importance.

[0078] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the specific implementation methods of the present invention will be described below with reference to the accompanying drawings. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings and other implementation methods can be obtained based on these drawings without creative work.

[0079] One embodiment of the present invention, as Figure 1 As shown, the present invention provides a vehicle-mounted non-contact contact line suspension state detection method, which is applied to a first area array camera and a second area array camera arranged in the direction of a suspension device of the contact line, comprising the steps of:

[0080] S100 controls the first area array camera to capture a first image at a preset time interval, and sends the first image to the vehicle-mounted industrial computer in real time.

[0081] Specifically, since the first picture taken by the first area array camera is used to identify the suspension state, the image accuracy requirement is relatively low. Therefore, the first area array camera can use a low-pixel area array camera to take the first picture, and adjust the preset time period of the first area array camera shooting according to the train form process, or use the first area array camera to shoot the video of the suspension state in real time to ensure the accuracy of suspension state recognition.

[0082] The S200 identifies whether there are key features of the suspension device in the first image through the vehicle-mounted industrial computer.

[0083] Specifically, the feature value of the first image is extracted by the on-board industrial computer, and the feature value of the first image is input into a preset first image recognition model, and the feature value is compared with the image containing the suspension device. If the output similarity is greater than a preset threshold, it is determined that the key features of the suspension device exist in the first image.

[0084] If yes, S300 controls the second area array camera to capture the suspension state of the contact line by the onboard industrial computer to obtain a second image, and controls the second area array camera to send the second image to the onboard industrial computer in real time.

[0085] Specifically, the second area array camera needs to be a high-precision, high-pixel area array camera. When the industrial computer determines that there are key features of the suspension device in the first image, the second area array camera is controlled in real time to capture images of the suspension state of the contact line at the current position.

[0086] S400 detects whether there is any abnormality in the suspension state of the contact line in the second picture through the on-board industrial computer.

[0087] Specifically, the feature value of the second image is extracted by the on-board industrial computer, and the feature value of the second image is input into a preset second image recognition model, and the feature value is compared with the image of the abnormal suspension state. If the output similarity is greater than a preset threshold, it is determined that the suspension state in the second image is abnormal.

[0088] The on-board non-contact contact wire suspension state detection method provided in the present embodiment sets two kinds of area array cameras. When the industrial computer determines that the first area array camera has photographed the suspension device, it controls the second area array camera to photograph the current suspension state of the contact wire, and specifically identifies whether the parts are intact based on the picture taken by the second area array camera. According to the shooting result of the first area array camera, it can be identified whether there is a suspension device at the current position, and there is no need to reserve a large amount of storage space to store the video data of the suspension state of the contact wire taken by the first area array camera, thereby improving the accuracy and convenience of detecting the on-board non-contact contact wire suspension state during train travel.

[0089] In one embodiment, Figure 2 As shown, step S200 uses the on-board industrial computer to identify whether there are key features of the suspension device in the first image, specifically including:

[0090] S211 identifies, by means of the vehicle-mounted industrial computer, whether there is a first straight line perpendicular to the x-axis of the image coordinate system, a second straight line horizontal to the x-axis of the image coordinate system, and a third straight line at an oblique angle to the x-axis of the image coordinate system in the first image.

[0091] Specifically, determine whether there are vertical straight lines, horizontal straight lines, and oblique straight lines in the first picture, where the vertical straight line represented by the first straight line corresponds to the column, the horizontal straight line represented by the second straight line corresponds to the horizontal arm, and the oblique straight line represented by the third straight line corresponds to the oblique wrist arm.

[0092] If the judgment result is yes in step S212, it is determined again whether the first image includes the first sub-feature and the second sub-feature.

[0093] Specifically, the first sub-feature is that the lengths of the first straight line, the second straight line and the third straight line are all within the corresponding preset length thresholds, and the second sub-feature is that there is an intersection between any two of the first straight line, the second straight line and the third straight line.

[0094] If the result of the second determination is yes, it is determined at S213 that the key features of the suspension device exist in the first image.

[0095] This embodiment discloses a method for identifying straight line features in a first image to determine whether there are key features of a suspension device in the first image, thereby balancing the accuracy and efficiency of identifying the suspension device in the first image and facilitating real-time control of a second area array camera to capture images.

[0096] In one embodiment, Figure 3 As shown, step S200 uses the on-board industrial computer to identify whether there are key features of the suspension device in the first image, and specifically includes:

[0097] S221 compares the first image with a plurality of image templates of contact wire suspension devices pre-stored in the vehicle-mounted industrial computer through the vehicle-mounted industrial computer, and calculates the image matching degree between the first image and each image template respectively.

[0098] S222: If there is a picture matching degree greater than a preset first matching degree threshold, it is determined that there is a key feature of the suspension device in the first picture.

[0099] Specifically, steps S221 to S222 and steps S211 to S213 can be respectively used as two implementation methods of step S200 to identify whether there are key features of the suspension device in the first picture through the vehicle-mounted industrial computer, or they can jointly determine whether there are key features of the suspension device in the first picture. The two schemes can be mutually verified to improve the accuracy of recognition.

[0100] In one embodiment, Figure 4 As shown, step S400 detects whether there is an abnormality in the suspension state of the contact line in the second picture by using the on-board industrial computer, which specifically includes:

[0101] S410 determines whether there is a fourth straight line perpendicular to the x-axis of the image coordinate system, a fifth straight line horizontal to the x-axis of the image coordinate system, and a sixth straight line at an oblique angle to the x-axis of the image coordinate system in the second image.

[0102] If the judgment result is yes, it is determined again at step S420 whether the second image includes the third sub-feature, the fourth sub-feature and the fifth sub-feature.

[0103] Specifically, the third sub-feature is that the lengths of the fourth straight line, the fifth straight line and the sixth straight line are all within the corresponding preset length thresholds, the fourth sub-feature is that the slopes of the fourth straight line, the fifth straight line and the sixth straight line are all within the corresponding preset slope thresholds, and the fifth sub-feature is that there is an intersection between any two straight lines of the fourth straight line, the fifth straight line and the sixth straight line.

[0104] If the result of the second judgment is yes, the intersection point between any two straight lines among the fourth straight line, the fifth straight line and the sixth straight line is taken as a key point, and an area of ​​a preset pixel area is intercepted in the second image with each key point as the center to generate a key point image.

[0105] S450 compares each key point image with the part data image stored in the database of the vehicle-mounted industrial computer to obtain a first comparison result.

[0106] Specifically, by pre-training a comparison neural network model based on part data images stored in a database of the vehicle-mounted industrial computer, the feature value of each key point image is extracted, and the feature value of each key point image is input into the comparison neural network model to obtain a first comparison result.

[0107] S460 calculates the part distances between the key points in the second image, and after standardizing the part distances, compares the part distances with the distance data stored in the database of the onboard industrial computer to obtain a second comparison result.

[0108] S470 determines whether there is an abnormality in the suspension state of the contact line according to the first comparison result and the second comparison result.

[0109] Specifically, if the first comparison result is that the similarity between each key point image and the part data image stored in the database of the vehicle-mounted industrial computer is greater than the preset similarity threshold, and the first comparison result is that the error between the calculated part distance and the distance data stored in the database is less than the preset error threshold, it is judged that there is no abnormality. If there is a comparison result greater than the preset similarity threshold or greater than the preset error threshold in the first comparison result and the second comparison result, it is judged that there is an abnormality.

[0110] The vehicle-mounted non-contact contact line suspension state detection method provided in this embodiment discloses a scheme for identifying whether there is an abnormality in the suspension state of the contact line in the second picture. By comparing the similarity between the key point picture and the part data picture in the second picture, as well as the part distance between the key points and the distance data stored in the database, it is judged whether the suspension state of the contact line is abnormal. The suspension state of the contact line is judged comprehensively based on multiple factors to improve the accuracy of detection.

[0111] In one embodiment, Figure 5As shown, if the result of the second judgment is yes in step S430, the intersection point between any two straight lines among the fourth straight line, the fifth straight line and the sixth straight line is taken as a key point, and after taking each key point as the center in the second picture, an area with a preset pixel area is intercepted to generate a key point picture, before step S450 compares each key point picture with the part data picture stored in the database of the vehicle-mounted industrial computer to obtain a first comparison result, it also includes:

[0112] S441 performs edge recognition on the key point image.

[0113] Specifically, edge recognition of key point images is performed through dilation, erosion, and grayscale gradient methods.

[0114] S442 filters the key point images according to the filtering features.

[0115] Specifically, the screening features include the area of ​​the key point image, the aspect ratio of the key point image, and the slope of the edge of the key point image. Key point images whose area exceeds the preset value, whose aspect ratio exceeds the limit, or whose edge slope does not match are eliminated. Subsequent operations are performed on key point images whose above three screening features are within the preset range.

[0116] The vehicle-mounted non-contact contact wire suspension state detection method provided in this embodiment screens key point images through edge recognition, thereby further improving the accuracy of contact wire suspension state detection. At the same time, it can avoid the impact of erroneous data on the accuracy of model training when the key point images are stored in the industrial computer for model training.

[0117] In one embodiment, Figure 6 As shown, after step S470 determines whether the suspension state of the contact line is abnormal according to the first comparison result and the second comparison result, it also includes:

[0118] If the judgment result is yes in S481, each key point image is compared with the part image data stored in the database of the vehicle-mounted industrial computer again to obtain a third comparison result, and it is determined whether each third comparison result is greater than a preset similarity threshold.

[0119] Specifically, it is determined again whether each key point image contains parts of the suspension device.

[0120] S482 If there is a third comparison result that is not greater than the similarity threshold, a parts missing alarm message is generated.

[0121] S483 If all third comparison results are greater than the similarity threshold, the second image is compared with the suspension state image stored in the database of the vehicle-mounted industrial computer to obtain a second matching degree, and the accuracy of the abnormal detection result of the suspension state of the contact line is verified according to the second matching degree.

[0122] Specifically, if the second matching degree is greater than the preset second matching degree threshold, the abnormal detection result of the suspension state of the contact line is updated, and it is determined that there is no abnormality in the suspension state of the contact line. The second image, the key point image corresponding to the second image, and the part distance corresponding to the second image are stored in the database of the on-board industrial computer to facilitate subsequent training and comparison of the neural network model based on this information. If the second matching degree is not greater than the second matching degree threshold, a missing parts alarm message is generated.

[0123] The vehicle-mounted non-contact contact wire suspension state detection method provided in the present embodiment compares the key point image with the part image again after determining that the contact wire suspension state is abnormal, so as to avoid calculation errors in the first comparison process. When the similarity between the key point image and the part image is greater than the similarity threshold, the second image is compared with the suspension state image stored in the database. Multiple verifications greatly improve the accuracy of the abnormality detection results.

[0124] In one embodiment, Figure 7 As shown, a constantly on LED 3 is provided to match the first area array camera 2 , and a flashing LED 4 is provided to match the second area array camera 1 .

[0125] In step S400, during the process of detecting whether there is an abnormality in the suspension state of the contact line in the second picture by the on-board industrial computer, the flashing LED is controlled to be in a working state at the same time, so as to avoid the flashing LED being always in a working state and affecting the safety of the train operation.

[0126] The first area array camera 2 and the second area array camera 1 are both arranged in two groups. The first area array camera 2 and the second area array camera 2 in the same group are placed side by side. The two groups of cameras are placed opposite to each other, so as to respectively take pictures and identify the two sides of the suspension device. An acceleration sensor 5 is also arranged for real-time detection of the acceleration of the train.

[0127] In one embodiment, Figure 8 As shown, the vehicle-mounted industrial computer receives the acceleration data sent by the acceleration sensor in real time. After step S400 detects whether the suspension state of the contact line in the second picture is abnormal through the vehicle-mounted industrial computer, the method further includes:

[0128] When S500 detects an abnormality, it controls the onboard industrial computer to calculate the current position of the train based on the acceleration data, and determines the number of the suspension device of the abnormal contact line based on the current position of the train.

[0129] The vehicle-mounted non-contact contact wire suspension state detection method provided in this embodiment receives the acceleration data sent by the acceleration sensor in real time. After detecting the abnormal suspension state, it determines the current position of the train according to the acceleration data, and promptly issues an early warning based on the location and number of the suspension device fault, so as to facilitate timely maintenance of the suspension device.

[0130] In one embodiment, Fig. 9 As shown, the present invention also provides a vehicle-mounted non-contact contact wire suspension state detection system, including a first area array camera 10, a vehicle-mounted industrial computer 20 and a second area array camera 30.

[0131] The first area array camera 10 is arranged toward the suspension device of the contact line, and is used to capture a first picture at preset time intervals, and send the first picture to the vehicle-mounted industrial computer 20 in real time. The vehicle-mounted industrial computer 20 is connected to the first area array camera 10, and is used to identify whether there are key features of the suspension device in the first picture. The second area array camera 30 is connected to the vehicle-mounted industrial computer 20, and is arranged toward the suspension device of the contact line. When the vehicle-mounted industrial computer 20 identifies the key features of the suspension device in the first picture, the second area array camera 30 captures the suspension state of the contact line according to the control of the vehicle-mounted industrial computer 20 to obtain a second picture, and sends the second picture to the vehicle-mounted industrial computer 20 in real time. The vehicle-mounted industrial computer 20 is also used to detect whether there is any abnormality in the suspension state of the contact line in the second picture after receiving the second picture.

[0132] Specifically, since the first picture taken by the first area array camera 10 is used to identify the suspension state, the image accuracy requirement is relatively low. Therefore, the first area array camera 10 can use a low-pixel area array camera to shoot the first picture, and adjust the preset time period of the first area array camera 10 according to the train form process, or use the first area array camera 10 to shoot a video of the suspension state in real time to ensure the accuracy of suspension state recognition.

[0133] The second area array camera 30 needs to be a high-precision and high-pixel area array camera. When the industrial computer determines that there are key features of the suspension device in the first image, the second area array camera 30 is controlled in real time to collect images of the suspension state of the contact line at the current position.

[0134] In the process of extracting the feature value of the first image through the on-board industrial computer 20, the on-board industrial computer 20 inputs the feature value of the first image into the preset first image recognition model, and compares the feature value with the image containing the suspension device. If the output similarity is greater than the preset threshold, it is judged that the key features of the suspension device exist in the first image. In the process of extracting the feature value of the second image through the on-board industrial computer 20, the on-board industrial computer 20 inputs the feature value of the second image into the preset second image recognition model, and compares the feature value with the image of the abnormal suspension state. If the output similarity is greater than the preset threshold, it is judged that the suspension state in the second image is abnormal.

[0135] The on-board non-contact contact line suspension state detection system provided in this embodiment is provided with two kinds of array cameras. When the industrial computer determines that the first array camera has photographed the suspension device, it controls the second array camera to photograph the current suspension state of the contact line, and specifically identifies whether the parts are intact based on the pictures taken by the second array camera. According to the photographing result of the first array camera, it can identify whether there is a suspension device at the current position, and there is no need to reserve a large amount of storage space to store the video data of the suspension state of the contact line photographed by the first array camera, thereby improving the accuracy and convenience of detecting the suspension state of the on-board non-contact contact line during the train driving process. In the above embodiments, the description of each embodiment has its own emphasis. For the part that is not described or recorded in detail in a certain embodiment, please refer to the relevant description of other embodiments.

[0136] Those of ordinary skill in the art will appreciate that the units and steps of each example described in conjunction with the embodiments disclosed herein can be implemented with electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed with hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0137] In the embodiments provided in the present application, it should be understood that the disclosed vehicle-mounted non-contact contact line suspension state detection method and system can be implemented in other ways. For example, the above-described vehicle-mounted non-contact contact line suspension state detection method and system embodiment is only schematic. For example, the division of the modules or units is only a logical function division. There may be other division methods in actual implementation. For example, multiple units or modules can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the communication connection between each other shown or discussed can be through some interfaces, communication connections of devices or units or integrated circuits, which can be electrical, mechanical or other forms.

[0138] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0139] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0140] It should be noted that the above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principle of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A vehicle-mounted non-contact contact wire suspension state detection method, characterized in that: The first area array camera and the second area array camera arranged in the direction of the suspension device facing the contact line include the following steps: Controlling the first area array camera to capture a first picture at a preset time interval, and sending the first picture to the vehicle-mounted industrial computer in real time; identifying, by the on-board industrial computer, whether the key features of the suspension device exist in the first image; If yes, the vehicle-mounted industrial computer controls the second area array camera to capture the suspension state of the contact line to obtain a second picture, and controls the second area array camera to send the second picture to the vehicle-mounted industrial computer in real time; The on-board industrial computer is used to detect whether there is any abnormality in the suspension state of the contact line in the second picture, including: extracting the feature value of the second picture by the on-board industrial computer, and inputting the feature value of the second picture into a preset second picture recognition model, and comparing the feature value with the picture of the abnormal suspension state. If the output similarity is greater than a preset threshold, it is determined that the suspension state in the second picture is abnormal.

2. A vehicle-mounted non-contact contact wire suspension state detection method according to claim 1, characterized in that: The identifying, by the on-board industrial computer, whether the key features of the suspension device exist in the first picture specifically includes: Identify, by the vehicle-mounted industrial computer, whether there is a first straight line perpendicular to the x-axis of the image coordinate system, a second straight line horizontal to the x-axis of the image coordinate system, and a third straight line at an oblique angle to the x-axis of the image coordinate system in the first image; If the judgment result is yes, secondly judge whether the first image includes a first sub-feature and a second sub-feature, the first sub-feature being that the lengths of the first straight line, the second straight line and the third straight line are all within the corresponding preset length threshold, and the second sub-feature being that there is an intersection between any two of the first straight line, the second straight line and the third straight line; If the secondary determination result is yes, it is determined that the key features of the suspension device exist in the first picture.

3. A vehicle-mounted non-contact contact wire suspension state detection method according to claim 1, characterized in that: The step of identifying, by the on-board industrial computer, whether the key features of the suspension device exist in the first image specifically includes: The vehicle-mounted industrial computer compares the first image with a plurality of image templates of the contact wire suspension device pre-stored in the vehicle-mounted industrial computer, and calculates the image matching degree between the first image and each of the image templates; If the picture matching degree is greater than a preset first matching degree threshold, it is determined that the key feature of the suspension device exists in the first picture.

4. A vehicle-mounted non-contact contact wire suspension state detection method according to claim 1, characterized in that: The detecting whether the suspension state of the contact line in the second picture is abnormal by the on-board industrial computer specifically includes: Determine whether there is a fourth straight line perpendicular to the x-axis of the image coordinate system, a fifth straight line horizontal to the x-axis of the image coordinate system, and a sixth straight line at an oblique angle to the x-axis of the image coordinate system in the second picture; If the judgment result is yes, secondly judge whether the second picture includes the third sub-feature, the fourth sub-feature and the fifth sub-feature, the third sub-feature is that the lengths of the fourth straight line, the fifth straight line and the sixth straight line are all within the corresponding preset length threshold, the fourth sub-feature is that the slopes of the fourth straight line, the fifth straight line and the sixth straight line are all within the corresponding preset slope threshold, and the fifth sub-feature is that there is an intersection between any two of the fourth straight line, the fifth straight line and the sixth straight line; If the secondary judgment result is yes, taking the intersection point between any two straight lines among the fourth straight line, the fifth straight line and the sixth straight line as the key point, and taking each of the key points as the center in the second picture, intercepting an area of ​​a preset pixel area to generate a key point picture; Comparing each of the key point images with the parts data images stored in the database of the vehicle-mounted industrial computer to obtain a first comparison result; Calculating the part distance between each of the key points in the second image, and after standardizing the part distance, comparing the part distance with the distance data stored in the database of the vehicle-mounted industrial computer to obtain a second comparison result; It is determined whether the suspension state of the contact line is abnormal according to the first comparison result and the second comparison result.

5. A vehicle-mounted non-contact contact wire suspension state detection method according to claim 4, characterized in that: After the region of the preset pixel area is intercepted to generate the key point image, and before the first comparison result is obtained by comparing each of the key point images with the part image data stored in the database of the vehicle-mounted industrial computer, the method further includes: Performing edge recognition on the key point image; The key point images are screened according to screening features, where the screening features include the area of ​​the key point images, the aspect ratio of the key point images, and the slope of the edge of the key point images.

6. A vehicle-mounted non-contact contact wire suspension state detection method according to any one of claims 4 to 5, characterized in that: After judging whether the suspension state of the contact line is abnormal according to the first comparison result and the second comparison result, the method further includes: If the judgment result is yes, again compare each of the key point images with the parts image data stored in the database of the vehicle-mounted industrial computer to obtain a third comparison result, and judge whether each of the third comparison results is greater than a preset similarity threshold; If there is the third comparison result that is not greater than the similarity threshold, generating a parts missing alarm message; If each of the third comparison results is greater than the similarity threshold, the second image is compared with the suspension state image stored in the database of the vehicle-mounted industrial computer to obtain a second matching degree, and the accuracy of the abnormal detection result of the suspension state of the contact line is verified based on the second matching degree.

7. A vehicle-mounted non-contact contact wire suspension state detection method according to claim 6, characterized in that: The step of verifying the accuracy of the abnormal detection result of the suspension state of the contact line according to the second matching degree specifically includes: If the second matching degree is greater than a preset second matching degree threshold, the abnormality detection result of the suspension state of the contact line is updated, it is determined that there is no abnormality in the suspension state of the contact line, and the second image, the key point image corresponding to the second image, and the part distance corresponding to the second image are stored in the database of the on-board industrial computer; If the second matching degree is not greater than the second matching degree threshold, the part missing alarm information is generated.

8. A vehicle-mounted non-contact contact wire suspension state detection method according to claim 1, characterized in that: Setting a constantly on LED that matches the first area array camera, and a flashing LED that matches the second area array camera; When the on-board industrial computer controls the second area array camera to capture the second picture, the flashing LED is controlled to be in working state at the same time.

9. A vehicle-mounted non-contact contact wire suspension state detection method according to any one of claims 1 to 5, 7 to 8, characterized in that: The vehicle-mounted industrial computer receives acceleration data sent by the acceleration sensor in real time; After the on-board industrial computer detects whether the suspension state of the contact line in the second picture is abnormal, the method further includes: When an abnormality is detected, the onboard industrial computer is controlled to calculate the current position of the train according to the acceleration data, and the number of the suspension device of the contact line where the abnormality occurs is determined according to the current position of the train.

10. A vehicle-mounted non-contact contact wire suspension state detection system, characterized in that: include: A first area array camera, which is arranged toward the suspension device of the contact line and is used to capture a first picture at a preset time interval and send the first picture to the vehicle-mounted industrial computer in real time; The vehicle-mounted industrial computer is connected to the first area array camera and is used to identify whether the key features of the suspension device exist in the first image; A second area array camera is connected to the vehicle-mounted industrial computer, and is arranged toward the suspension device of the contact wire. The second area array camera is used to capture the suspension state of the contact wire to obtain a second image under the control of the vehicle-mounted industrial computer when the vehicle-mounted industrial computer recognizes that there are key features of the suspension device in the first image, and send the second image to the vehicle-mounted industrial computer in real time. The vehicle-mounted industrial computer is further configured to detect whether there is any abnormality in the suspension state of the contact line in the second image after receiving the second image.

Images