A station computer interlocking turnout status information detection system and method

Abstract

The present invention relates to a station computer interlocking turnout status information detection system and method thereof, the system comprises a video acquisition terminal and a data processing unit, the video acquisition terminal is set up at a position where the station computer interlocking display screen interface can be clearly photographed, and the interlocking display screen station field image is non-contactly acquired; the data processing unit is connected to the video acquisition terminal, and the turnout symbol image is extracted from the original frame image, and the turnout symbol image is scaled and image processed, and the matching coefficient between the grayscale symbol of the turnout to be detected and the grayscale symbol of the standard turnout is calculated to obtain the turnout status information detection result. Compared with the prior art, the present invention can achieve the purpose of real-time detection of station computer interlocking turnout status information in a non-intrusive manner without contact, and can reduce the detection engineering workload and improve the detection safety.

Description

Technical Field

[0001] The invention relates to the technical field of rail transit, and in particular to a station computer interlocking turnout state information detection system and method thereof. Background Art

[0002] The computer interlocking system is responsible for the operation of the railway station route and is the core control equipment for railway operation safety. The computer interlocking system realizes the interlocking control between the turnouts, signal machines and track circuits in the station under the operation of the signal operator or the ATS system. It is an indispensable guarantee equipment for safe and efficient railway operation. The computer interlocking system has been widely used in the world. In areas with more developed railway transportation, the computer interlocking system is no longer an isolated signal control system, but a subsystem of the signal safety comprehensive monitoring system and the comprehensive operation management system.

[0003] The station computer interlocking system is a functional safety system, and its connection with other systems is strictly restricted. In the railway transportation system, the train depot, engineering depot, vehicle depot and other units all need to obtain the station computer interlocking information to quickly understand the real-time status of the train and station yard, which is conducive to the development of work. In order to obtain the turnout status information, the station computer interlocking system is often required to provide a one-way electrically isolated communication interface, and cooperate with the preparation of corresponding special communication protocols to achieve the acquisition of computer interlocking information. The whole process is time-consuming and very complicated. In addition, since the station computer interlocking information needs to be obtained by invading the communication interface, there is a high possibility of communication security risks. Summary of the invention

[0004] The purpose of the present invention is to overcome the defects of the above-mentioned prior art and provide a station computer interlocking turnout status information detection system and method thereof, so as to achieve the purpose of quickly and safely detecting the station computer interlocking turnout status information.

[0005] The object of the present invention can be achieved by the following technical scheme: a station computer interlocking turnout status information detection system, characterized in that it includes a video acquisition terminal and a data processing unit, the video acquisition terminal is set up at a position where the station computer interlocking display screen station screen can be clearly photographed, the video acquisition terminal is connected to the data processing unit, the video acquisition terminal is used to obtain the station screen image of the station computer interlocking display screen (3); the data processing unit is used to analyze and process the station screen image and perform matching calculation to obtain the corresponding detection result.

[0006] Furthermore, the data processing unit includes a system initialization module, an icon extraction module, a pixel scaling module, an image processing module, a matching calculation module and a detection judgment module:

[0007] The system initialization module is used to calibrate the station screen image and mark the turnout icon information;

[0008] The icon extraction module is used to extract the corresponding turnout icon to be detected from each frame of real-time image according to the calibrated turnout icon information;

[0009] The pixel scaling module is used to scale the image of the symbol to be detected to have the same number of pixels as the standard symbol image;

[0010] The image processing module is used to convert the icon image into a clear grayscale image;

[0011] The matching calculation module is used to calculate the matching coefficient between the grayscale symbol to be detected and the standard grayscale symbol;

[0012] The detection and judgment module analyzes and judges the data output by the matching calculation module to output corresponding detection results.

[0013] A station computer interlocking turnout status information detection method comprises the following steps:

[0014] S1, the video acquisition terminal is installed at a fixed position to shoot the station computer interlocking display screen station screen, the specific steps include the camera installed at the fixed position to shoot the video, the video is transmitted to the data processing unit through the communication network, the frame image is extracted from the video and stored, and recorded as the original frame image;

[0015] S2. Using any frame image obtained in step S1 as a sample image, the system initialization module completes system initialization under manual intervention and stores relevant initialization information in the data processing unit. The system initialization includes station field image calibration, station field image extraction, and information calibration of turnout icons in station field image;

[0016] S3. According to the station computer interlocking display specification, the image processing module converts the standard turnout symbols of various turnout states into corresponding grayscale images in advance to obtain standard turnout grayscale symbols. The turnout states include four states: positioning, reverse position, loss of indication, and squeezed turnout;

[0017] S4, using the initialization information obtained in step S2, the icon extraction module extracts all the turnout icons to be detected from each frame of the received real-time image;

[0018] S5, converting all the turnout symbols to be detected into corresponding grayscale turnout symbols to be detected;

[0019] S6, the matching calculation module calculates the matching coefficients of all grayscale symbols of turnouts to be detected and the grayscale symbols of standard turnouts;

[0020] S7. The detection and judgment module analyzes and judges the status detection results of each turnout to be detected based on the calculation results of each matching coefficient.

[0021] Furthermore, the specific steps of system initialization in step S2 are:

[0022] S21, the station scene image calibration is to calibrate the original frame image Frame into a horizontal calibration image Frame′, so that the station scene image in the calibration image is rectangular, so as to improve the accuracy of subsequent detection;

[0023] The specific process of step S21 is as follows:

[0024] S211, transform the coordinates of each pixel point in the original frame image Frame into the corresponding coordinates in the calibration image Frame′ by using the transformation calculation formula, and the transformation calculation formula is as follows:

[0025]

[0026]

[0027] Where (x, y) is the coordinate of the pixel in the original image, and (x′, y′) is the corresponding coordinate after transformation;

[0028] S212, deform the transformation calculation formula, and obtain the coordinates of the four vertices of the station scene image in the original frame image Frame by manual calibration [(x1, y1), (x2, y2), (x3, y3), (x4, y4)], assume that the corresponding coordinates of the four vertices in the calibration image Frame' are [(x'1, y'1), (x'2, y'1), (x'1, y'2), (x'2, y'2)], substitute the coordinates of the four groups of vertices before and after the transformation, and obtain the following matrix:

[0029]

[0030] By solving the above matrix, we can obtain the parameter a required for the station image calibration: i , (i=1, 2, ..., 7, 8), storing the parameters in a data processing unit for use in subsequent steps;

[0031] S22, the station field image extraction is to extract the station field image area R in the calibration image Frame', and the rectangular frame parameters of the station field image area R are the four vertex coordinates [(x'1, y'1), (x'2, y'1), (x'1, y'2), (x'2, y'2)] in step S212, where x'1<x'2, y'1<y'2, and the area R is defined as:

[0032] R∈(x, y), x′1≤x≤x′2, y′1≤y≤y′2

[0033] Take the region R from the image Frame′, set the pixel point with coordinates (x′1, y′1) as the coordinate origin, and obtain a station screen image D with a size of (M1×M2), where M1=x′2-x′1, M2=y′2-y′1, and save the vertex coordinates [(x′1, y′1), (x′2, y′1), (x′1, y′2), (x′2, y′2)] to the data processing unit for use in subsequent steps;

[0034] S23, the information calibration of the turnout symbol in the station yard image is to obtain the information of all the turnout symbols to be detected in the station yard image D by manual calibration. The turnout symbol is the smallest graphic unit used to represent the turnout. The turnout information includes the turnout number C i , fork tip coordinates (x i ,y i ) and turnout direction i , where i = 1, 2, ..., n, n is the number of turnouts in the station image D, and the turnout number C i is the number of the i-th turnout in the system, and the turnout point coordinate (x i ,y i ) is the coordinate of the pixel point of the i-th turnout in the station image D, and the turnout direction Dire i is the direction of the i-th turnout point, with a value of left / right. The turnout information is saved to the data processing unit for use in subsequent steps.

[0035] Furthermore, the specific steps of extracting the turnout icon to be detected in step S4 are:

[0036] S41, using the parameter a obtained in step S21 and step S22 i , (i=1, 2, ..., 7, 8) and vertex coordinates [(x′1, y′1), (x′2, y′1), (x′1, y′2), (x′2, y′2)], the original frame image Frame γ Calibrate to Frame′ γ , and extract the station scene image D from it γ , where γ is the frame number;

[0037] S42, using the turnout information obtained in step S23, in the station screen image D γ In the figure, the fork tip coordinate (x i ,y i ) as the center, extract the square R with side length L i , the vertex coordinates of the square are defined as And according to the direction of the turnout i , change the direction of the turnout i The right square R i Flip left and right to get the turnout symbol P to be detected γi , number the turnout C i And the corresponding turnout symbol P to be detected γi Stored in the data processing unit for use in subsequent steps.

[0038] Furthermore, the specific calculation formula for converting the grayscale image in step S3 and step S5 is:

[0039]

[0040] Among them, z is the grayscale image after the symbol image is converted, and the matrix is the color image corresponding to the symbol, o mn is the pixel corresponding to the mth row and nth column in the color image of the symbol, and They represent the components of the symbol image in the three color channels of R, G, and B respectively. Parameters A1, A3, and A3 are the weight coefficients corresponding to each color component. The weight coefficients are determined based on experience to make the turnout image and the background have a larger grayscale difference.

[0041] Furthermore, the step S5 specifically includes the following steps:

[0042] Step S51: scaling all the turnout icons to be detected to have the same number of pixels as the standard turnout icon;

[0043] Step S52: converting all the scaled image of the symbol to be detected into corresponding grayscale images respectively;

[0044] Step S53: bilaterally filter the grayscale image to remove noise and obtain the grayscale symbol P of the turnout to be detected. gγi .

[0045] Furthermore, the calculation formula of the matching coefficient in step S6 is specifically:

[0046]

[0047] N=m×n

[0048] Among them, S γij is the grayscale symbol P of the i-th turnout to be detected in the γ-th frame image gγi and the jth standard turnout grayscale symbol P sj The matching coefficient between them, N is the number of pixels of the symbol, m is the height, n is the width, P gγikis the kth pixel of the i-th turnout grayscale symbol to be detected in the γ-th frame image, P sjk is the kth pixel of the jth standard turnout grayscale symbol

[0049] Furthermore, the step S7 specifically includes the following steps:

[0050] S71, select the matching coefficient S with the largest value from the matching coefficients of the grayscale symbol of the turnout to be detected and the four types of standard turnout grayscale symbols. γij ;

[0051] S72, determine the matching coefficient S γij Is it greater than or equal to the threshold S? m If the answer is yes, the detection is successful, and the state represented by the standard icon corresponding to the maximum matching coefficient is obtained, that is, the γth frame image number C is output. i The current state of the turnout is j; if the judgment is no, that is, the matching coefficients of the grayscale symbol of the turnout to be detected and the standard grayscale symbols of the four types of turnout states do not reach the threshold S m , the turnout to be tested does not belong to one of the four defined states.

[0052] Compared with the prior art, the present invention has the following advantages:

[0053] 1. The present invention sets a video acquisition terminal and a data processing unit, uses the video acquisition terminal to obtain the station field picture of the station computer interlocking display screen, and uses the data processing unit to analyze and process the real-time acquired station field picture image and perform matching calculations to obtain the corresponding detection results, thereby achieving the purpose of real-time non-contact and non-invasive detection of the station computer interlocking turnout status information, effectively reducing the detection engineering workload, improving the detection speed, and ensuring the detection safety.

[0054] 2. The data processing unit in the present invention can accurately and quickly extract the turnout symbol to be detected from the collected image through the pre-calibrated station screen calibration information and turnout symbol information, and then convert the standard symbol image of each turnout state into a corresponding grayscale image based on the display specification or the display rules defined by each subway company to obtain the standard turnout grayscale symbol. Finally, the accuracy of the detection result can be guaranteed by converting the grayscale image of the turnout symbol to be detected, calculating and analyzing the matching coefficients of all the turnout grayscale symbols to be detected and the standard turnout symbol grayscale images. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] Figure 1 It is a schematic diagram of the system structure of the present invention;

[0056] Figure 2 It is a schematic diagram of the method flow of the present invention;

[0057] Figure 3a is the original frame image in the embodiment;

[0058] Figure 3b is a calibration image in the embodiment;

[0059] Figure 3c The station scene image extracted in the embodiment;

[0060] Figure 4 The grayscale icon of the standard turnout in the embodiment;

[0061] Figure 5a The graying result of the turnout symbol to be detected in the embodiment;

[0062] Figure 5b is the grayscale symbol of the turnout to be detected in the embodiment;

[0063] Explanation of the markings in the figure: 1. Video acquisition terminal, 2. Data processing unit, 3. Computer interlocking display screen. DETAILED DESCRIPTION

[0064] The present invention is described in detail below with reference to the accompanying drawings and specific embodiments.

[0065] Example

[0066] like Figure 1 As shown, a station computer interlocking turnout status information detection system includes a video acquisition terminal 1 and a data processing unit 2. The video acquisition terminal 1 is set up at a position where the station computer interlocking display screen 3 can be clearly photographed. The video acquisition terminal 1 is connected to the data processing unit 2. In this embodiment, the video acquisition terminal 1 is connected to the data processing unit 2 via a data transmission cable. The video acquisition terminal 1 is used to obtain the station screen image of the station computer interlocking display screen 3; the data processing unit 2 is used to analyze and process the operation interface image and perform matching calculations to obtain the corresponding detection results.

[0067] The data processing unit 2 includes a system initialization module, a symbol extraction module, a pixel scaling module, an image processing module, a matching calculation module and a detection judgment module.

[0068] The system initialization module is used to calibrate the station screen image and mark the turnout icon information;

[0069] The icon extraction module is used to extract the corresponding turnout icon to be detected from each frame of real-time image according to the calibrated turnout icon information;

[0070] The pixel scaling module is used to scale the image of the symbol to be detected to the same number of pixels as the standard symbol image;

[0071] The image processing module is used to convert the icon image into a clear grayscale image;

[0072] The matching calculation module is used to calculate the matching coefficient between the grayscale symbol to be detected and the standard grayscale symbol;

[0073] The detection and judgment module analyzes and judges the data output by the matching calculation module to output the corresponding detection results.

[0074] In addition, in order to facilitate the user to directly obtain the test results, the data processing unit 2 is also provided with a display screen for displaying the test results.

[0075] The above system is applied in practice to realize a station computer interlocking turnout status information detection method, such as Figure 2 As shown, the following steps are included:

[0076] S1, the video acquisition terminal is installed at a fixed position to shoot the station computer interlocking display screen station screen, the specific steps include the camera installed at the fixed position to shoot the video, the video is transmitted to the data processing unit through the communication network, the frame image is extracted from the video and stored, and recorded as the original frame image;

[0077] S2. Using any frame image obtained in step S1 as a sample image, the system initialization module completes system initialization under manual intervention and stores relevant initialization information in the data processing unit. The system initialization includes station field image calibration, station field image extraction, and information calibration of turnout icons in station field image;

[0078] S3. According to the station computer interlocking display specification, the image processing module converts the standard turnout symbols of various turnout states into corresponding grayscale images in advance to obtain standard turnout grayscale symbols. The turnout states include four states: positioning, reverse position, loss of indication, and squeezed turnout;

[0079] S4, using the initialization information obtained in step S2, the icon extraction module extracts all the turnout icons to be detected from each frame of the received real-time image;

[0080] S5, converting all the turnout symbols to be detected into corresponding grayscale turnout symbols to be detected;

[0081] S6, the matching calculation module calculates the matching coefficients of all grayscale symbols of turnouts to be detected and the grayscale symbols of standard turnouts;

[0082] S7. The detection and judgment module analyzes and judges the status detection results of each turnout to be detected based on the calculation results of each matching coefficient.

[0083] This embodiment applies the above technical solution, and its specific process includes:

[0084] B100: The video acquisition terminal is installed at a fixed position to capture the station field image of the station computer interlocking display screen. The specific steps include capturing video with a camera installed at a fixed position, transmitting the video to a data processing unit through a communication network, extracting frame images from the video and storing them as original frame images.

[0085] B200: Take a frame image obtained in B100 as a sample image, such as Figure 3a As shown, the system initialization module completes the system initialization under manual intervention and stores the relevant initialization information in the data processing unit. The system initialization includes station field image calibration, station field image extraction, and information calibration of the turnout icon in the station field image, including the following steps:

[0086] B2001: Calibrate the original frame image Frame to the horizontal calibration image Frame′, so that the station scene image in the calibration image is rectangular. The original frame image Frame is as follows: Figure 3a As shown, the calibration image Frame′ is Figure 3b As shown, the specific steps are:

[0087] B20011: transform the coordinates of each pixel point in the original frame image Frame into the corresponding coordinates in the calibration image Frame′ by the transformation calculation formula. The transformation calculation formula is as follows:

[0088]

[0089]

[0090] Where (x, y) is the coordinate of the pixel in the original image, and (x′, y′) is the corresponding coordinate after transformation;

[0091] B20012: The transformation calculation formula is deformed, and the coordinates of the four vertices of the station scene image in the original frame image Frame are obtained by manual calibration [(x1, y1), (x2, y2), (x3, y3), (x4, y4)], and the corresponding coordinates of the four vertices in the calibration image Frame' are assumed to be [(x'1, y'1), (x'2, y'1), (x'1, y'2), (x'2, y'2)], and the coordinates of the four groups of vertices before and after the transformation are substituted to obtain the following matrix:

[0092]

[0093] By solving the above matrix, we can obtain the parameter a required for station image calibration: i , (i=1,2,…,7,8).

[0094] In this embodiment, the coordinates of the four vertices in the original frame image Frame are [(450, 126), (5501, 658), (372, 1244), (5556, 1446)], and the corresponding coordinates of the four vertices in the calibration image Frame′ are [(450, 120), (5500, 120), (450, 1300), (5500, 1300)].

[0095] The calibration parameter a calculated from this i for:

[0096] (6.5187e-1, 5.7700e-2, 1.3851e+2, -1.1826e-1, 1.0570e0, 3.7134e+1, -6.0181e-5, 2.2964e-5)

[0097] Then the calibration parameter a i Stored in the data processing unit for use in subsequent steps.

[0098] B2002: Extract the station scene image region R in the calibration image Frame′. The rectangular frame parameters of the station scene image region R are the four vertex coordinates [(x′1, y′1), (x′2, y′1), (x′1, y′2), (x′2, y′2)] in step S212, where x′1<x′2, y′1<y′2. The region R is defined as:

[0099] R∈(x, y), x′1≤x≤x′2, y′1≤y≤y′2

[0100] Take the region R from the image Frame′, set the pixel with coordinates (x′1, y′1) as the origin, and obtain the station scene image D with a size of (M1×M2), where M1=x′2-x′1, M2=y′2-y′1, as shown in Figure 3c shown.

[0101] In this embodiment, the coordinates of the four vertices are [(450, 120), (5500, 120), (450, 1300), (5500, 1300)]. The coordinates of the four vertices are stored in the data processing unit for use in subsequent steps.

[0102] B2003: In the station image D, the information of all the turnout symbols to be detected is obtained by manual calibration. The turnout symbol refers to the smallest graphic unit representing the turnout. The turnout information includes the turnout number C i , fork tip coordinates (x i ,y i ) and turnout direction i, where i = 1, 2, ..., n, n is the number of turnouts in the station image D, and the turnout number C i is the number of the i-th turnout in the system, and the turnout point coordinate (x i ,y i ) is the coordinate of the pixel point of the i-th turnout in the station image D, and the turnout direction Dire i The direction of the i-th turnout point is left / right, and the turnout information is saved to the data processing unit for use in subsequent steps.

[0103] B300: According to the station computer interlocking display specification, the image processing module converts the standard turnout symbols of various turnout states into corresponding grayscale images in advance to obtain the standard turnout grayscale symbols, such as Figure 4 As shown in the figure, the turnout status includes four types: positioning, reverse position, loss of indication, and squeezed turnout, which are denoted by P sj , j=1,2,3,4.

[0104] B400: For each frame of real-time image, the symbol extraction module uses the initialization information obtained in B200 to extract all the turnout symbols to be detected and record them as P γi , where γ is the frame number, the specific steps are:

[0105] B4001: Use the parameter a obtained in steps B2001 and B2002 i , (i=1, 2, ..., 7, 8) and vertex coordinates [(x′1, y′1), (x′2, y′1), (x′1, y′2), (x′2, y′2)], the original frame image Frame γ Calibrate to Frame′ γ , and extract the station scene image D from it γ , where γ is the frame number;

[0106] B4002: Use the turnout information obtained in step B2003 to display the turnout information in the station image D. γ In the figure, the fork tip coordinate (x i ,y i ) as the center, extract the square R with side length L i , the vertex coordinates of the square are defined as And according to the direction of the turnout i , change the direction of the turnout i The right square R i Flip left and right to get the turnout symbol P to be detected γi , number the turnout C i And the corresponding turnout symbol P to be detected γi Stored in the data processing unit for use in subsequent steps.

[0107] In this embodiment, the turnout number C i = 1301, the turnout point coordinates (x i ,y i ) is (952, 465), the turnout direction Dire i For the right, the side length L of the square is set to 100, and the square R is obtained i , the square R i After flipping left and right, the turnout symbol P to be detected is obtained. γi .

[0108] B500: Set all turnout icons to be tested P γi are respectively converted into the corresponding grayscale symbol P of the turnout to be detected gγi , the specific steps include:

[0109] B5001: Scale all the turnout symbols to be detected to the same number of pixels as the standard turnout symbol;

[0110] B5002: Convert all the scaled images of the symbols to be detected into corresponding grayscale images, such as Figure 5a As shown, the conversion process can be expressed as the following formula:

[0111]

[0112] Among them, z is the grayscale image after the symbol image is converted, and the matrix is the color image corresponding to the symbol, o mn is the pixel corresponding to the mth row and nth column in the color image of the symbol, and Respectively represent the components of the symbol image in the three color channels of R, G, and B. Parameters A1, A3, and A3 are weight coefficients corresponding to the color components, which are 0.299, 0.587, and 0.114, respectively, in this embodiment;

[0113] B5003: Perform bilateral filtering on the grayscale image to remove noise and obtain the grayscale symbol P of the turnout to be detected. gγi ,like Figure 5b shown.

[0114] B600: The matching calculation module calculates the matching coefficients of all grayscale symbols of the turnouts to be detected and the grayscale symbols of the standard turnouts. The detection judgment module analyzes and judges the state detection results corresponding to the turnout symbols to be detected according to the calculation results of the matching coefficients.

[0115] B700: The data processing unit repeats B400 to B600, outputs the status of each switch in the frame, and repeats this cycle to realize the real-time detection function.

[0116] The grayscale conversion process in step B300 is specifically as follows:

[0117] B3001: According to the image size of the standard turnout symbol (height m pixels, width n pixels), the standard turnout symbol image is converted from an RGB color image to a standard turnout grayscale symbol. The conversion process can be expressed as the following formula:

[0118]

[0119] Among them, z is the grayscale image after the symbol image is converted, and the matrix is the color image corresponding to the symbol, o mn is the pixel corresponding to the mth row and nth column in the color image of the symbol, and Respectively represent the components of the symbol image in the three color channels of R, G, and B. Parameters A1, A3, and A3 are weight coefficients corresponding to the color components, which are determined to be 0.299, 0.587, and 0.114 in this embodiment;

[0120] B3002: For all state standard turnout icons, execute B3001 in sequence, and finally obtain the standard turnout grayscale icons of all states {P s1 , P s2 , P s3 , P s4}, where 1, 2, 3, and 4 represent the four states of positioning, reverse positioning, loss of representation, and squeeze.

[0121] Step B600 specifically includes:

[0122] B6001: Calculate the first turnout symbol P to be detected in the γth frame gγ1 All standard turnout grayscale icons {P s1 , P s2 , P s3 , P s4}, the result can be recorded as the set {S γ11 , S γ12 , S γ13 , S γ13}, the following formula describes the grayscale image P of the symbol to be tested gγ1 The first standard symbol grayscale image P s1 The calculation method of the matching coefficient S1 is:

[0123]

[0124] N=m×n

[0125] Wherein, N is the number of image pixels, that is, N = m × n;

[0126] B6002: The matching coefficient set obtained in B6001 {S γ11 , S γ12 , S γ13 , S γ14} to find the largest value S γ1j , j = 1, 2, 3, 4, if it exceeds the specified threshold 0.8, then the output of the state of the C1 turnout in the γth frame is j. γ11 , S γ12 , S γ13 , S γ13} are all less than the specified threshold value (i.e. the turnout symbol to be detected is not similar to all the standard turnout symbols), then the output is "the turnout to be detected does not belong to one of the four defined states".

[0127] In this embodiment, Figure 5b The obtained matching coefficient set is {0.9, 0.6, 0.7, 0.7}, that is, the maximum value S γ11 =0.9>0.8, then the output state numbered 1301 in the γth frame is positioning.

[0128] B6003: For the grayscale image of the remaining turnouts to be detected in the frame image {P gγ2 , P gγ2 ,…,P gγn}, repeat steps B6001 and B6002 to determine the status of all the symbols to be tested in the frame image in turn.

[0129] In actual application, each time the camera takes a frame of real-time image, the data processing unit repeats the above steps to output the detection results of all the turnout icons to be detected in each frame of image, and the cycle is repeated to realize the real-time detection function. If the state of the detected turnout changes, the data processing unit will also record and save the turnout number, time, state before change and state after change when the state changes (this operation is still performed when the state changes to "not belonging to one of the four defined states").

[0130] In summary, the present technical solution does not rely on electrical connection with the station computer interlocking system, and can detect and obtain real-time station computer interlocking turnout status information in a non-contact, non-invasive manner without the need for direct electrical connection.

[0131] The above specific implementations do not constitute a limitation on the protection scope of the present invention. Without departing from the spirit and scope of the present invention, the present invention can be implemented in other specific forms (such as when the detection object is a display screen station screen of other train operation control related systems such as ATS and CTC). In addition, those skilled in the art can make various changes and deformations to the present invention without departing from the spirit and scope of the present invention, and these improvements and deformations should also be regarded as the protection scope of the present invention. Therefore, the attached claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.

Claims

1. A station computer interlocking turnout status information detection method, applied to a station computer interlocking turnout status information detection system, characterized in that: The detection system comprises a video acquisition terminal (1) and a data processing unit (2); the video acquisition terminal (1) is set up at a position where a station field interface of a station computer interlocking display screen (3) can be clearly photographed; the video acquisition terminal (1) is connected to the data processing unit (2); the video acquisition terminal (1) is used to obtain a station field screen image of the station computer interlocking display screen (3); the data processing unit (2) is used to analyze and process the station field screen image and perform matching calculations to obtain corresponding detection results; The data processing unit (2) comprises a system initialization module, a symbol extraction module, a pixel scaling module, an image processing module, a matching calculation module and a detection and judgment module. The system initialization module is used to calibrate the station image and mark the turnout symbol information. The icon extraction module is used to extract the corresponding turnout icon to be detected from each frame of real-time image according to the calibrated turnout icon information; The pixel scaling module is used to scale the image of the symbol to be detected to have the same number of pixels as the standard symbol image; The image processing module is used to convert the icon image into a clear grayscale image; The matching calculation module is used to calculate the matching coefficient between the grayscale symbol to be detected and the standard grayscale symbol; The detection and judgment module analyzes and judges the data output by the matching calculation module to output corresponding detection results; The detection method comprises the following steps: S1, the video acquisition terminal captures the video of the station computer interlocking display screen and transmits it to the data processing unit, which extracts and stores the frame image from the video and records it as the original frame image; S2, taking any original frame image as a sample image, performing initialization processing on the sample image, and storing relevant initialization information in a data processing unit, wherein the initialization processing process includes station field image calibration, station field image extraction, and information calibration of turnout icons in the station field image; S3. According to the technical specifications for the station computer interlocking operation display, the data processing unit converts the standard turnout symbols of various turnout states into corresponding grayscale images in advance to obtain standard turnout grayscale symbols. The turnout states include four states: positioning, reverse position, loss of indication, and squeezed turnout. S4, based on the initialization information in step S2, the data processing unit extracts all the turnout icons to be detected from each frame of the received real-time image; S5, the data processing unit converts all the turnout symbols to be detected into corresponding grayscale turnout symbols to be detected; S6, the data processing unit calculates the matching coefficients of all grayscale symbols of the turnouts to be detected and the grayscale symbols of the standard turnouts; S7. According to the calculation results of each matching coefficient, the data processing unit analyzes and determines to obtain the state detection result corresponding to each turnout to be detected.

2. A station computer interlocking turnout status information detection method according to claim 1, characterized in that: The specific process of the initialization process in step S2 is: S21, station screen image calibration: calibrate the original frame image Frame to the head-on calibration image Frame ′ , so that the station field image in the calibration image is rectangular to improve the accuracy of subsequent detection. The specific steps are: First, the coordinates of each pixel in the original frame image Frame are transformed into the coordinates of each pixel in the calibration image Frame through the transformation calculation formula. ′ The corresponding coordinates in the transformation calculation formula are as follows: Among them, (x, y) is the coordinate of the pixel point in the original image, (x ′ ,y ′ ) are the corresponding coordinates after transformation; Then, the transformation calculation formula is deformed, and the coordinates of the four vertices of the station scene image in the original frame image Frame are obtained by manual calibration [(x1, y1), (x2, y2), (x3, y3), (x4, y4)], and the calibration image Frame is set ′ The corresponding coordinates of the four vertices are [(x ′ 1,y ′ 1),(x ′ 2,y ′ 1),(x ′ 1,y ′ 2),(x ′ 2,y ′ 2)], substitute the coordinates of the four groups of vertices before and after the transformation, and get the following matrix: By solving the above matrix, we can obtain the parameter a required for station image calibration: i ,i=1,2,…,7,8, store the parameter in the data processing unit; S22, station screen image extraction: in the calibration image frame ′ In step S21, the station scene image region R is extracted, and the rectangular frame parameters of the station scene image region R are the four vertex coordinates [(x ′ 1,y ′ 1),(x ′ 2,y ′ 1),(x ′ 1,y ′ 2),(x ′ 2,y ′ 2)], where x ′ 1 <x ′ 2,y ′ 1 <y ′ 2. The region R is defined as: R∈(x,y),x ′ 1≤x≤x ′ 2,and ′ 1≤y≤y ′ 2 The region R is transformed from the image Frame ′ Take it out and take the coordinate (x ′ 1,y ′ 1) is set as the coordinate origin, and the station scene image D with a size of (M1×M2) is obtained, where M1=x ′ 2-x ′ 1,M2=y ′ 2-y ′ 1; The vertex coordinates [(x ′ 1,y ′ 1),(x ′ 2,y ′ 1),(x ′ 1,y ′ 2),(x ′ 2,y ′ 2)] is saved to the data processing unit; S23, information calibration of turnout icons in the station yard image: in the station yard image D, information of all turnout icons to be detected is obtained by manual calibration. The turnout icon is the smallest graphic unit representing the turnout. The information of the turnout icon includes the turnout number C i , fork tip coordinates (x i ,y i ) and turnout direction i , where i = 1, 2, ..., n, n is the number of turnouts in the station image D, and the turnout number C i is the number of the i-th turnout in the system, and the turnout point coordinate (x i ,y i ) is the coordinate of the pixel point of the i-th turnout in the station image D, and the turnout direction Dire i The direction of the i-th turnout point is left / right, and the turnout information is saved to the data processing unit.

3. A station computer interlocking turnout status information detection method according to claim 2, characterized in that: The step S4 specifically comprises the following steps: S41, using the parameter a obtained in step S21 and step S22 i , i=1,2,…,7,8 and vertex coordinates [(x ′ 1,y ′ 1),(x ′ 2,y ′ 1),(x ′ 1,y ′ 2),(x ′ 2,y ′ 2)], the original frame image Frame γ Calibrate to Frame ′ γ , and extract the station scene image D from it γ , where γ is the frame number; S42, using the turnout information obtained in step S23, in the station screen image D γ In the figure, the fork tip coordinate (x i ,y i ) as the center, extract the square R with side length L i , the vertex coordinates of the square are defined as And according to the direction of the turnout i , if the turnout direction Dire i If it is right, the extracted square R i Flip left and right to get the turnout symbol P to be detected γi , number the turnout C i And the corresponding turnout symbol P to be detected γi Stored in the data processing unit.

4. A station computer interlocking turnout status information detection method according to claim 3, characterized in that: The specific calculation formula for converting the grayscale image in step S3 and step S5 is: Among them, z is the grayscale image after the symbol image is converted, and the matrix is the color image corresponding to the symbol, o mn is the pixel corresponding to the mth row and nth column in the color image of the symbol, and They represent the components of the symbol image in the three color channels of R, G, and B respectively. Parameters A1, A3, and A3 are the weight coefficients corresponding to each color component. The weight coefficients are determined based on experience to make the turnout image and the background have a larger grayscale difference.

5. A station computer interlocking turnout status information detection method according to claim 1, characterized in that: The step S5 specifically comprises the following steps: S51, scaling all the turnout symbols to be detected to have the same number of pixels as the standard turnout symbol; S52, converting all the scaled image symbols to be detected into corresponding grayscale images respectively; S53, bilateral filtering is performed on the grayscale image to remove noise, and a grayscale symbol of the turnout to be detected is obtained.

6. A station computer interlocking turnout status information detection method according to any one of claims 1 to 5, characterized in that: The calculation formula of the matching coefficient in step S6 is specifically: N=m×n Among them, S γij is the grayscale symbol P of the i-th turnout to be detected in the γ-th frame image gγi and the jth standard turnout grayscale symbol P sj The matching coefficient between them, N is the number of pixels of the symbol, m is the height, n is the width, P gγik is the kth pixel of the i-th turnout grayscale symbol to be detected in the γ-th frame image, P sjk is the kth pixel of the jth standard turnout grayscale symbol.

7. A station computer interlocking turnout status information detection method according to claim 6, characterized in that: The step S7 specifically includes the following steps: S71, select the matching coefficient S with the largest value from the matching coefficients of the grayscale symbol of the turnout to be detected and the four types of standard turnout grayscale symbols. γij ; S72, determine the matching coefficient S γij Is it greater than or equal to the threshold S? m If the answer is yes, the detection is successful, and the state represented by the standard icon corresponding to the maximum matching coefficient is obtained, that is, the γth frame image number C is output. i The current state of the turnout is j; If the judgment is no, it means that the matching coefficients of the grayscale symbol of the turnout to be detected and the standard grayscale symbols of the four types of turnout states do not reach the threshold S. m , that is, the turnout to be detected does not belong to one of the four defined turnout states.

Images