A method for extracting and identifying track areas in complex turnout areas based on image recognition

Abstract

The present invention belongs to the field of smart urban rail and computer vision technology, and specifically proposes a method for extracting and identifying track areas in complex turnout areas based on image recognition. The method processes the input image through an instance segmentation model to obtain an n-dimensional matrix mask Mask; processes the matrix mask through a track area processing module to obtain a matrix mask Mask_Re; analyzes the number of track areas in Mask_Re, ​​and intercepts the turnout area image through a turnout area image acquisition module; obtains the turnout's fixed and reverse position state, that is, the train's direction of travel, through a target positioning model, thereby obtaining the unique track area in which the train is running; and performs visualization processing to output an RGB image result. The present invention processes complex turnout area scene images through an instance segmentation model, and locates and identifies the turnout tip state through a target positioning model. It can accurately identify and extract the track area in which the train is currently traveling from complex turnout area images obtained from different scenes and shooting angles.

Description

Technical Field

[0001] The present invention belongs to the field of autonomous driving and computer vision technology, and specifically relates to a method for extracting and identifying track areas in complex switch areas based on image recognition. Background Art

[0002] With the development of autonomous driving technology, the application of assisted driving technology in rail transit is becoming increasingly important. The core of assisted driving technology is to accurately identify the train's travel area, namely the "track zone." Based on this identification, defining the vehicle spacing limit is a key step in assisted driving technology in rail transit. Given the rapid development of the rail transit industry, the operating environment of national railways, subways, and trams has become increasingly complex. Multiple rails intersect at switch areas. Therefore, ensuring that trains can correctly identify the track zone in complex switch areas and improve driving safety is a critical issue that needs to be addressed.

[0003] At present, based on the track switch recognition technology, it is possible to assess possible safety risks in advance. However, when faced with switch scenarios with large changes in lighting differences and complex track intersections, the use of traditional track switch recognition technology has the problem of low recognition accuracy.

[0004] To address this issue, those skilled in the art have made numerous efforts. For example, Chinese patent application CN 111860137A proposes a vision-based track switch recognition method. This method utilizes pure vision to identify track switches, which can obtain color information and image distribution, thereby improving the accuracy of switch recognition. However, the recognition accuracy of switch images obtained in different scenarios is still insufficient. Summary of the Invention

[0005] To solve the above problems, the present invention provides a method for extracting and identifying track areas in complex turnout areas based on image recognition. The scene images of complex turnout areas are processed by an instance segmentation model, and are positioned and identified by a target positioning model. It can accurately identify and extract the track areas where trains are traveling in turnout images obtained from different scenes and shooting angles.

[0006] In order to achieve the above object, the technical solution adopted by the present invention is

[0007] A method for extracting and identifying track areas in a complex turnout area based on image recognition, the method comprising the following steps:

[0008] Step S1: Constructing a track area extraction and recognition system in a complex turnout area based on image recognition; including an instance segmentation model, a target positioning model, a track area processing module, and a visualization processing module;

[0009] Step S2: Obtain an image of the track ahead of the train, scale the image to R*C, and input the image into the instance segmentation model for processing to obtain an n-dimensional matrix mask, where n is the number of track areas.

[0010] Step S3: Analyze the number of track areas; if the number of track areas is equal to 1, input the mask matrix into the visualization module, and output the image result of unique track area identification; if the number of track areas is greater than 1, input the mask matrix into the track area processing module;

[0011] Step S4: The track area processing module includes an invalid track area filtering module and a switch area image acquisition module; the mask matrix is ​​processed by the invalid track area filtering module to obtain a processed track area image matrix Mask_Re;

[0012] Step S5: Analyze the number of track areas in Mask_Re. If the number of track areas is 1, input the Mask_Re matrix into the visualization module to output the image result of unique track area recognition. If the number of track areas is greater than 1, input Mask_Re into the switch area image acquisition module to intercept the switch area image.

[0013] Step S6: marking the turnout tips in the turnout branch area in the turnout area image into two types: open and close, as the turnout tip dataset;

[0014] Step S7: Using the turnout tip dataset as the training set and test set of the target positioning model, repeatedly training and testing are performed to obtain the final model;

[0015] Step S8: Using the turnout area image as input to the final model, the turnout tip is located and identified using the final model to obtain the turnout direction of the current turnout branch area, and further obtain the track area of ​​the train running in the turnout area;

[0016] Step S9: Visualize the acquired track area through a visualization module and output an RGB image result.

[0017] Furthermore, in step S2, the size of the n-dimensional matrix mask is the same as the size of the input image, the area with a value of 0 in the matrix is ​​a non-track area, the area with a value other than 0 is a track area, and the value corresponding to each track area is different.

[0018] Furthermore, in step S4, the mask matrix is ​​processed by filtering the invalid track area module to obtain the processed track area image matrix Mask_Re. The specific method is:

[0019] Step S4-1: Analyze the distribution of the mask matrix values ​​and calculate the coordinates of the center point of each track area when the vertical coordinate is 10 (Rail_x, 10);

[0020] Step S4-2: Calculate the difference between the horizontal coordinate of the center point of each track area when the vertical coordinate is 10 and the horizontal coordinate of the center point of the input image, |Rail_x-R / 2|; obtain the track area Mask_N with the smallest distance from the center point of the image; where R / 2 is the horizontal coordinate of the center point of the input image;

[0021] Step S4-3: Extract the track area that overlaps with Mask_N to obtain the processed track area image matrix Mask_Re.

[0022] Furthermore, in step S5, Mask_Re is input into the module for obtaining the switch area image. The specific method for intercepting the switch area image is as follows:

[0023] Calculate all overlapping areas, select the overlapping area A with the largest number of track areas, calculate the minimum circumscribed rectangle coordinates, and intercept the turnout area image based on the obtained rectangle coordinates.

[0024] Furthermore, in step S8, the switch tip is located and identified using the final model to obtain the switch direction of the current switch branch area, and then the specific method for obtaining the unique track area in the switch area is as follows:

[0025] Step S8-1: By locating and identifying the switch tip, the center coordinates (Sx, Sy) of the switch tip and the state of the switch tip: open or closed are obtained;

[0026] Step S8-2: Determine the direction of the turnout based on the coordinates of the upper left corner (Lx, Ly), the upper right corner (Rx, Ry), the open / closed state of the turnout tip, and the center coordinates (Sx, Sy) of the turnout tip in the minimum circumscribed rectangle of the turnout image.

[0027] Step S8-3: Select and extract the track area where the train runs from the overlapping area according to the direction of the turnout, and obtain the track area where the train runs in the turnout area.

[0028] Furthermore, in step S8-2, the specific method for determining the direction of the turnout is: when Sx is less than (Lx+Rx) / 2 and the state of the turnout tip is open, the turnout direction is left; when Sx is greater than (Lx+Rx) / 2 and the state of the turnout tip is open, the turnout direction is right.

[0029] Furthermore, in step S8-3, the specific method for selecting and extracting the track area for train operation from the overlapping area according to the direction of the turnout is as follows: when the turnout is left-opening, the track area on the left side of area A is searched within the minimum circumscribed rectangle of the overlapping area A and extracted; when the turnout is right-opening, the track area on the right side of A is searched within the minimum circumscribed rectangle of the overlapping area A and extracted.

[0030] The technical solution of the present invention has the following beneficial effects:

[0031] The present invention processes the input image through an instance segmentation model to obtain an n-dimensional matrix representing the track area, and filters the matrix through a track area processing module to process invalid track areas and obtain switch area images to obtain the intercepted switch area image, and then uses a target positioning model to combine the coordinate parameters of the upper left corner point, the coordinate parameters of the upper right corner point, the open and closed state of the switch tip and the coordinate parameters of the center point of the switch tip to judge the direction of the switch, and combines the minimum bounding rectangle range of the overlapping area to identify and extract the track area where the train is traveling, thereby realizing accurate identification and extraction of the track area where the train is traveling in the switch area in different scenarios, thereby improving the accuracy of the recognition process; at the same time, it realizes pixel-level segmentation and instance distinction of each track area in the scenario where there are multiple track areas in the switch area, with better generalization and robustness. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 This is a flow chart of a method for extracting and identifying track areas in a complex turnout area based on image recognition according to the present invention;

[0033] Figure 2 is a processing flow chart of the track area processing module of the present invention;

[0034] Figure 3 is a diagram of a turnout area captured in an embodiment of the present invention;

[0035] Figure 4 is a flow chart of the segmentation process according to an embodiment of the present invention;

[0036] Figure 5 This is a diagram showing the effect of track area extraction and recognition in an embodiment of the present invention. DETAILED DESCRIPTION

[0037] The technical solution of the present invention is further described below with reference to the accompanying drawings, but is not limited thereto. Any modification or equivalent replacement of the technical solution of the present invention that does not depart from the spirit and scope of the technical solution of the present invention should be included in the scope of protection of the present invention.

[0038] like Figure 1 and Figure 2 As shown, this embodiment proposes a method for extracting and identifying track areas in a complex turnout area based on image recognition, comprising the following steps:

[0039] Step S1: Construct a track area extraction and recognition system in a complex turnout area based on image recognition; including an instance segmentation model, a target positioning model, a track area processing module and a visualization processing module.

[0040] The DeepLab v3+ model used in the instance segmentation model is used to normalize the input image to match the training data distribution and obtain an n-dimensional matrix consisting of multiple track areas; Figure 4 As shown in the figure, the instance segmentation model architecture design includes an encoder and a decoder. The encoder is based on the ResNet backbone network, gradually extracting high-level semantic features through multiple residual blocks and dilated convolutional layers. ResNet consists of a 7x7 convolutional layer, a max pooling layer, and four residual blocks. The last two residual blocks use dilated convolutions to expand the receptive field. At the end of the encoder, the Spatial Pyramid Pooling (ASPP) module captures multi-scale information through parallel 1x1 convolutional layers, three 3x3 dilated convolutional layers with different dilation rates, and a global average pooling layer. The output feature maps of all branches are reduced in dimensionality through a 1x1 convolutional layer and concatenated together, and then further reduced in dimensionality through another 1x1 convolutional layer. The decoder uses bilinear interpolation to upsample the low-resolution feature maps output by the ASPP module to the same size as the input image. It then extracts low-level feature maps from the intermediate layers of the backbone network and fuses them with the upsampled high-level feature maps through element-by-element addition to restore fine-grained spatial information. Finally, a 1x1 convolutional layer outputs a per-pixel category prediction. Depthwise separable convolutions are used extensively throughout the network to reduce computational cost, and batch normalization layers are added after each convolutional layer to accelerate training and improve model generalization.

[0041] The processing process is:

[0042] Use convolutional neural networks (CNNs) to extract deep features of the image; these features will be used in subsequent region proposal networks (RPNs) and mask predictions;

[0043] Slide a small convolution kernel over the feature map to generate a series of potential bounding box proposals and assign an objectness score to each proposal;

[0044] Regress the proposed bounding boxes to obtain more accurate locations;

[0045] Map the bounding box output by RPN to the feature map, and extract fixed-size feature map patches through the region of interest (RoI) pooling operation;

[0046] Each RoI pooled feature block is processed by two parallel fully connected layers: one for classification, that is, to determine whether it is an object; the other for generating a mask;

[0047] Based on the output of the mask branch, a binary mask is generated for each proposed region. The generated mask is upsampled using a deconvolution layer to obtain finer mask edges. Non-maximum suppression (NMS) is performed on masks with higher classification scores to remove overlapping masks and ensure that each object is segmented only once. The NMS-processed mask is output as the final segmentation result, where each mask corresponds to an instance in the image, that is, each track region.

[0048] Among them, the track area processing module is used to filter out invalid track areas from the obtained matrix and obtain turnout area images.

[0049] Among them, the target positioning model is used to locate and identify the switch tip in the acquired switch area image to obtain the center coordinates of the switch tip and the opening and closing status of the switch tip; and the opening direction of the switch is determined by combining the coordinates of the upper left corner and the upper right corner points of the minimum circumscribed rectangle in the acquired switch area image, so as to accurately extract the track area where the train is traveling.

[0050] Among them, the visualization module is used to visualize the track area and convert it into an RGB image.

[0051] Step S2: Obtain an image of the track ahead of the train, scale the image to R*C, and feed the image into the instance segmentation model for processing to obtain an n-dimensional matrix mask, where n is the number of track sections. Specifically, the size of the n-dimensional matrix mask is the same as the input image size. Regions with a value of 0 in the matrix are non-track sections, and regions with a non-zero value are track sections. Each track section has a different value.

[0052] For example, when there are three track areas, the mask matrix is ​​3D, with the first dimension being Mask_1. Areas with a value of 0 in Mask_1 correspond to non-track areas, while areas with a value of 1 correspond to track area instance 1. Take the 3D vector in the xth row and yth column of the mask matrix. If the number of non-zero values ​​in the vector is greater than or equal to 2, it indicates that the track areas overlap. The pixel with the xth coordinate and the yth coordinate in the original input image is the overlapping area of ​​the track areas.

[0053] Step S3: Analyze the number of track areas; if the number of track areas is equal to 1, input the mask matrix into the visualization module and output the image result of unique track area identification; if the number of track areas is greater than 1, input the mask matrix into the track area processing module.

[0054] Step S4: The track area processing module includes an invalid track area filtering module and a switch area image acquisition module; the mask matrix is ​​processed by the invalid track area filtering module to obtain the processed track area image matrix Mask_Re. The specific method is:

[0055] Step S4-1: Analyze the distribution of the mask matrix values ​​and calculate the coordinates of the center point of each track area when the vertical coordinate is 10 (Rail_x, 10);

[0056] Step S4-2: Calculate the difference between the horizontal coordinate of the center point of each track area when the vertical coordinate is 10 and the horizontal coordinate of the center point of the input image, |Rail_x-R / 2|; obtain the track area Mask_N with the smallest distance from the center point of the image; where R / 2 is the horizontal coordinate of the center point of the input image;

[0057] Step S4-3: Extract the track area that overlaps with Mask_N to obtain the processed track area image matrix Mask_Re.

[0058] Step S5: Analyze the number of track areas in Mask_Re. If the number of track areas is 1, input the Mask_Re matrix into the visualization module, perform visualization processing on the matrix, and output the image result of unique track area identification; if the number of track areas is greater than 1, input Mask_Re into the switch area image acquisition module to intercept the switch area image. The specific method is:

[0059] like Figure 3 As shown, all overlapping areas are calculated, and the overlapping area A with the largest number of track areas is selected. For example, the n-dimensional vector of the x-th row and y-th column in the mask matrix is ​​taken, and different values ​​other than 0 are extracted to form a list. The set of pixels with the longest list length is the required overlapping area A. Then, the minimum circumscribed rectangle coordinates are calculated, and the switch area image is intercepted according to the obtained rectangular coordinates.

[0060] Specifically, the formula for calculating the coordinates of the minimum bounding rectangle is as follows. Let the point set be , the coordinates of the lower left corner and upper right corner of the minimum bounding rectangle are:

[0061] lower left corner

[0062] upper right corner

[0063] Step S6: marking the turnout tips in the turnout branch area in the turnout area image into two types: open and close, as the turnout tip dataset.

[0064] Step S7: The turnout point dataset is used as the training and test sets for the target localization model, and repeated training and testing are performed to obtain the final model. Specifically, the target localization model, also known as the target detection model, uses the Yolov5 model, which divides the input image into an S×S grid. Each grid cell is responsible for detecting targets whose center point falls within that cell. Each grid allows the prediction of two bounding boxes, which can be understood as S×S×2 prediction areas covering the entire area of ​​the image. Object detection is then performed within these prediction areas. The model uses Darknet53 as the backbone network, consisting of 53 convolutional layers. The network begins with a standard convolutional layer, followed by a batch normalization layer and a leaky ReLU activation function, followed by multiple residual blocks. Each residual block contains multiple convolutional layers and is optimized by applying the Cross Stage Partial (CSP) network strategy.

[0065] Step S8: Using the image of the turnout area as the input of the final model, the turnout tip is located and identified using the final model to obtain the turnout direction of the current turnout branch area, and then the track area where the train is running in the turnout area. The specific method is as follows:

[0066] Step S8-1: By locating and identifying the switch tip, the center coordinates (Sx, Sy) of the switch tip and the state of the switch tip: open or closed are obtained;

[0067] Step S8-2: Determine the direction of the turnout based on the coordinates of the upper left corner (Lx, Ly), the coordinates of the upper right corner (Rx, Ry), the open / closed state of the turnout tip, and the center coordinates (Sx, Sy) of the turnout tip in the minimum circumscribed rectangle of the turnout image. When Sx is less than (Lx+Rx) / 2 and the turnout tip is open, the turnout direction is left; when Sx is greater than (Lx+Rx) / 2 and the turnout tip is open, the turnout direction is right.

[0068] Step S8-3: Select and extract the track area for train operation from the overlapping area according to the direction of the turnout, and obtain the track area for train operation in the turnout area. When the turnout is left-opening, search for and extract the track area on the left side of area A within the minimum circumscribed rectangle of the overlapping area A; when the turnout is right-opening, search for and extract the track area on the right side of A within the minimum circumscribed rectangle of the overlapping area A.

[0069] Step S9: Visualize the acquired track area through a visualization module and output an RGB image result.

[0070] In order to verify the reliability and effectiveness of the method proposed in the present invention, the method proposed in the present invention was applied to a practical scenario, experimental verification was carried out, and the experimental results were analyzed.

[0071] In the verification experiment, a camera is used as the image input. A camera with an appropriate focal length is installed in front of the train driver's cab, and the camera focal length is fine-tuned so that the image can clearly display the track area. The software uses the camera structure function to obtain the current image in real time, and the obtained image is used as the input image for experimental verification.

[0072] The method for extracting and identifying track areas in complex turnout areas based on image recognition described in the present invention can extract and identify turnouts in different scenarios, thereby accurately identifying the track area that the train needs to travel, and realizing pixel-level segmentation and instance differentiation of each track area in the scenario where there are multiple track areas in the turnout area, with better generalization and robustness. Figure 5 This is a diagram showing the effect of track area extraction and recognition in an embodiment of the present invention, where (a) is the original image, (b) is the result of instance segmentation model processing, (c) is the extraction result of the track area processing module, and (d) is the RGB image result output by visualization processing.

[0073] The above descriptions are merely embodiments of the present invention and are not intended to limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made using the contents of the present invention's description and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A method for extracting and identifying track areas in complex turnout areas based on image recognition, characterized in that: The following steps are involved: Step S1: Constructing a track area extraction and recognition system in a complex turnout area based on image recognition; It includes an instance segmentation model, a target positioning model, a track area processing module and a visualization module; wherein the track area processing module includes an invalid track area filtering module and a switch area image acquisition module; Step S2: Obtain an image of the track ahead of the train, scale the image to R*C, and input the image into the instance segmentation model for processing to obtain an n-dimensional matrix mask, where n is the number of track areas; Step S3: Analyze the number of track areas; if the number of track areas is equal to 1, input the mask matrix into the visualization module and output the image result of unique track area identification; if the number of track areas is greater than 1, input the mask matrix into the track area processing module; Step S4: Processing the mask matrix by the invalid track area filtering module in the track area processing module to obtain a processed track area image matrix Mask_Re; Step S5: Analyze the number of track areas in Mask_Re. If the number of track areas is 1, input the Mask_Re matrix into the visualization module to output an image result of unique track area identification. If the number of track areas is greater than 1, input Mask_Re into the acquisition module of the turnout area image in the track area processing module to intercept the turnout area image. Step S6: marking the turnout tips in the turnout branch area in the turnout area image into two types: open and close, as the turnout tip dataset; Step S7: using the turnout tip dataset as a training set and a test set for the target positioning model, repeatedly training and testing to obtain a final model; Step S8: Using the turnout area image as input to the final model, the turnout tip is located and identified using the final model to obtain the turnout direction of the current turnout branch area, and further obtain the track area of ​​the train running in the turnout area; Step S9: Visualize the acquired track area through the visualization module and output an RGB image result.

2. The method for extracting and identifying track areas in complex turnout areas based on image recognition according to claim 1, characterized in that: In step S2, the size of the n-dimensional matrix mask is the same as the size of the input image. The area with a value of 0 in the matrix is ​​a non-track area, and the area with a value other than 0 is a track area. The value corresponding to each track area is different.

3. The method for extracting and identifying track areas in complex turnout areas based on image recognition according to claim 1, characterized in that: In step S2, the DeepLab v3+ model used in the instance segmentation model first uses a convolutional neural network to extract deep features of the image; slides a small convolution kernel on the feature map to generate a series of potential bounding box proposals and assigns an objectness score to each proposal; regresses the proposed bounding boxes to obtain more accurate locations; maps the bounding boxes output by the RPN to the feature map, and extracts fixed-size feature blocks through the region of interest (RoI) pooling operation; each RoI-pooled feature block is processed by two parallel fully connected layers: one for classification, that is, to determine whether it is an object; and the other for mask generation; based on the output of the mask branch, a binary mask is generated for each proposed region; the generated mask is upsampled using a deconvolution layer to obtain finer mask edges; non-maximum suppression (NMS) is performed on masks with higher classification scores to remove overlapping masks to ensure that each object is segmented only once; the mask after NMS processing is output as the final segmentation result, and each mask corresponds to an instance in the image; The model consists of an encoder and a decoder; the encoder part is based on the ResNet backbone network, and gradually extracts high-level semantic features through multiple residual blocks and void convolution layers. ResNet contains a 7x7 convolution layer, a maximum pooling layer and four residual blocks. The last two residual blocks use void convolution to expand the receptive field; at the end of the encoder, the spatial pyramid pooling (ASPP) module captures multi-scale information through parallel 1x1 convolution layers, three 3x3 void convolution layers with different void rates and a global average pooling layer. The output feature maps of all branches are reduced in dimension and spliced ​​together through a 1x1 convolution layer, and then further reduced in dimension through a 1x1 convolution layer; the decoder part uses bilinear interpolation to upsample the low-resolution feature map output by the ASPP module to the same size as the input image, and extracts low-level feature maps from the middle layer of the backbone network, and fuses them with the upsampled high-level feature maps by element-by-element addition. Finally, a 1x1 convolution layer outputs the category prediction for each pixel.

4. The method for extracting and identifying track areas in complex turnout areas based on image recognition according to claim 1, characterized in that: In step S4, the mask matrix is ​​processed by filtering the invalid track area module to obtain the processed track area image matrix Mask_Re as follows: Step S4-1: Analyze the distribution of the mask matrix values ​​and calculate the coordinates of the center point of each track area when the vertical coordinate is 10 (Rail_x, 10); Step S4-2: Calculate the difference between the horizontal coordinate of the center point of each track area when the vertical coordinate is 10 and the horizontal coordinate of the center point of the input image, |Rail_x-R / 2|; obtain the track area Mask_N with the smallest distance from the center point of the image; where R / 2 is the horizontal coordinate of the center point of the input image; Step S4-3: Extract the track area that overlaps with Mask_N to obtain the processed track area image matrix Mask_Re.

5. The method for extracting and identifying track areas in complex turnout areas based on image recognition according to claim 1, characterized in that: In step S5, Mask_Re is input into the module for obtaining the image of the turnout area. The specific method for intercepting the image of the turnout area is as follows: Calculate all overlapping areas, select the overlapping area A with the largest number of track areas, calculate the minimum circumscribed rectangle coordinates, and intercept the turnout area image based on the obtained rectangle coordinates.

6. The method for extracting and identifying track areas in complex turnout areas based on image recognition according to claim 1, characterized in that: In step S7: The target localization model is also known as the target detection model. The YOLOv5 model used in this target detection model divides the input image into an S×S grid. Each grid cell is responsible for detecting the target whose center point falls within the cell. Each grid is allowed to predict two bounding boxes, which can be understood as S×S×2 prediction areas, covering the entire area of ​​the image. Target detection is then performed in these prediction areas. The model uses Darknet53 as the backbone network, which consists of 53 convolutional layers. The network starts with a normal convolutional layer, followed by a batch normalization layer and a Leaky ReLU activation function, and then consists of multiple residual blocks. Each residual block contains multiple convolutional layers and is optimized by applying the Cross Stage Partial (CSP) network strategy.

7. The method for extracting and identifying track areas in complex turnout areas based on image recognition according to claim 1, characterized in that: In step S8, the final model is used to locate and identify the turnout tip, obtain the turnout direction of the current turnout branch area, and then obtain the unique track area in the turnout area. The specific method is: Step S8-1: By locating and identifying the switch tip, the center coordinates (Sx, Sy) of the switch tip and the state of the switch tip: open or closed are obtained; Step S8-2: Determine the direction of the turnout based on the coordinates of the upper left corner (Lx, Ly), the upper right corner (Rx, Ry), the open / closed state of the turnout tip, and the center coordinates (Sx, Sy) of the turnout tip in the minimum circumscribed rectangle of the turnout image. Step S8-3: Select and extract the track area where the train runs from the overlapping area according to the direction of the turnout, and obtain the track area where the train runs in the turnout area.

8. The method for extracting and identifying track areas in complex turnout areas based on image recognition according to claim 7, characterized in that: In step S8-2, the specific method for determining the direction of the turnout is: when Sx is less than (Lx+Rx) / 2 and the state of the turnout tip is open, the turnout direction is left; when Sx is greater than (Lx+Rx) / 2 and the state of the turnout tip is open, the turnout direction is right.

9. The method for extracting and identifying track areas in complex turnout areas based on image recognition according to claim 7, characterized in that: In step S8-3, the specific method for selecting and extracting the track area for train operation from the overlapping area according to the direction of the turnout is as follows: when the turnout is left-opening, the track area on the left side of area A is searched within the minimum circumscribed rectangle of the overlapping area A and extracted; when the turnout is right-opening, the track area on the right side of A is searched within the minimum circumscribed rectangle of the overlapping area A and extracted.

Images