Method for determining position of battery car in shield tunnel

Abstract

The invention discloses a method for determining the position of a battery car in a shield tunnel, and relates to the technical field of image processing, and the method comprises the steps: carrying out the image collection of the interior of the shield tunnel in real time through a collection device carried on a target battery car of which the position is to be determined, and obtaining a first image corresponding to the current moment; performing feature extraction on the first image through a semantic segmentation algorithm, extracting coding features in the first image, and identifying the coding features to obtain a numeric symbol identification result corresponding to the coding features; and determining position information corresponding to the target battery car of which the position is to be determined according to the numerical symbol identification result. Based on the digital OCR recognition algorithm, digital detection is achieved, and therefore the current accurate position of the battery car is obtained based on the tunnel side wall number.

Description

Technical Field

[0001] This invention relates to the field of image processing technology, and in particular to a method for determining the location of a battery-powered vehicle inside a shield tunnel. Background Technology

[0002] Currently, electric vehicles used in shield tunneling scenarios lack real-time positioning capabilities. Drivers primarily rely on observation or intuition to estimate their position, confirming it with workers at the shield machine via intercom upon approaching. Furthermore, positioning within the tunnel mainly relies on beacons and SLAM algorithms. The beacon method requires deploying a beacon transmitter at regular intervals within the tunnel, each marked with location information, which is then read by a reading device on the electric vehicle. However, as the tunnel advances, continuous beacon deployment is necessary, resulting in high costs.

[0003] SLAM-related positioning algorithms also rely on a fixed environment and the construction of high-precision maps. As the shield tunnel continues to advance, positioning cannot be achieved by building a map once, and continuous mapping is required, which is quite difficult to implement in actual engineering.

[0004] The positioning accuracy requirements for shield tunneling battery vehicles are not high, but low cost and minimal repetitive work are required, so a more convenient and cost-effective method is urgently needed. Summary of the Invention

[0005] The technical problem to be solved by this invention is to address the shortcomings of existing technologies, specifically by providing a method for determining the location of a battery-powered vehicle inside a shield tunnel, as detailed below:

[0006] 1) In a first aspect, the present invention provides a method for determining the location of a battery-powered vehicle inside a shield tunnel, the specific technical solution of which is as follows:

[0007] By using the acquisition device mounted on the target battery vehicle at the location to be determined, images of the inside of the shield tunnel are acquired in real time to obtain the first image corresponding to the current moment.

[0008] The first image is subjected to feature extraction using a semantic segmentation algorithm. The encoded features in the first image are extracted, and the encoded features are identified to obtain the digital symbol recognition result corresponding to the encoded features.

[0009] Based on the digital symbol recognition results, the location information corresponding to the target electric vehicle at the location to be determined is determined.

[0010] The beneficial effects of the method for determining the location of a battery-powered vehicle in a shield tunnel provided by this invention are as follows:

[0011] First, by using a data acquisition device mounted on an electric vehicle to collect images of the tunnel interior in real time, detailed information about the tunnel's interior can be quickly obtained without being limited by the time and frequency of manual inspection, greatly improving inspection efficiency. Second, semantic segmentation algorithms are used to extract features from the images. Compared to traditional methods, semantic segmentation algorithms can automatically extract higher-level semantic features and more accurately identify coded features in the images, effectively avoiding errors caused by manual identification. Third, the location information of the electric vehicle can be determined based on the digital symbol recognition results, enabling precise positioning of the electric vehicle and facilitating better management and scheduling, thereby improving the efficiency and safety of tunnel construction and maintenance.

[0012] Based on the above solution, the present invention can be further improved as follows.

[0013] Furthermore, the semantic segmentation algorithm performs downsampling on the input data by combining max pooling, average pooling, and convolution operations with a Concat network.

[0014] Furthermore, the process of identifying the encoded features is as follows:

[0015] The encoded features are identified using a digital OCR recognition algorithm;

[0016] The digital OCR recognition algorithm includes: convolutional layers, bidirectional LSTM, and fully connected layers.

[0017] Furthermore, the process of determining the location information corresponding to the target electric vehicle at the location to be determined is as follows:

[0018] By using a pre-defined table of correspondence between target digital symbol recognition results and target location information, the target location information corresponding to the digital symbol recognition results is determined, and the target location information is identified as the location information corresponding to the target electric vehicle at the location to be determined.

[0019] 2) Secondly, the present invention also provides a system for determining the location of a battery-powered vehicle inside a shield tunnel, the specific technical solution of which is as follows:

[0020] The acquisition module is used to: acquire images of the inside of the shield tunnel in real time through an acquisition device mounted on a target battery vehicle at the location to be determined, and obtain the first image corresponding to the current moment;

[0021] The recognition module is used to: extract features from the first image using a semantic segmentation algorithm, extract coded features from the first image, recognize the coded features, and obtain the recognition result of the digital symbol corresponding to the coded features;

[0022] The determining module is used to: determine the location information corresponding to the target electric vehicle at the location to be determined based on the digital symbol recognition result.

[0023] Based on the above solution, the present invention can be further improved as follows.

[0024] Furthermore, the semantic segmentation algorithm performs downsampling on the input data by combining max pooling, average pooling, and convolution operations with a Concat network.

[0025] Furthermore, the process of identifying the encoded features is as follows:

[0026] The encoded features are identified using a digital OCR recognition algorithm;

[0027] The digital OCR recognition algorithm includes: convolutional layers, bidirectional LSTM, and fully connected layers.

[0028] Furthermore, the process of determining the location information corresponding to the target electric vehicle at the location to be determined is as follows:

[0029] By using a pre-defined table of correspondence between target digital symbol recognition results and target location information, the target location information corresponding to the digital symbol recognition results is determined, and the target location information is identified as the location information corresponding to the target electric vehicle at the location to be determined.

[0030] 3) In a third aspect, the present invention also provides an electronic device, the electronic device including a processor coupled to a memory, the memory storing at least one computer program, the at least one computer program being loaded and executed by the processor to enable the electronic device to perform any of the above methods.

[0031] 4) In a fourth aspect, the present invention also provides a computer-readable storage medium storing at least one computer program, which is loaded and executed by a processor to enable a computer to implement any of the above methods.

[0032] It should be noted that the beneficial effects of the technical solutions of the second to fourth aspects of the present invention and their corresponding possible implementations can be found in the above description of the technical effects of the first aspect and its corresponding possible implementations, and will not be repeated here. Attached Figure Description

[0033] Other features, objects, and advantages of the invention will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings:

[0034] Figure 1 This is a flowchart illustrating a method for determining the location of a battery-powered vehicle inside a shield tunnel, according to an embodiment of the present invention.

[0035] Figure 2This is a tunnel sidewall coding semantic segmentation structure diagram of a method for determining the position of a battery-powered vehicle in a shield tunnel according to an embodiment of the present invention.

[0036] Figure 3 This is a schematic diagram of the downsampling layer structure of a method for determining the position of a battery-powered vehicle in a shield tunnel according to an embodiment of the present invention.

[0037] Figure 4 This is a tunnel sidewall coding and identification network structure diagram of a method for determining the position of a battery-powered vehicle in a shield tunnel, according to an embodiment of the present invention.

[0038] Figure 5 This is a structural framework diagram of an electronic device according to the present invention. Detailed Implementation

[0039] To make the objectives, technical solutions and advantages of the present invention more clear, the embodiments of the present invention will be described in further detail below with reference to the accompanying drawings.

[0040] like Figure 1 As shown in the figure, a method for determining the location of a battery-powered vehicle in a shield tunnel according to an embodiment of the present invention includes the following steps:

[0041] S1, by using the acquisition device mounted on the target battery vehicle at the location to be determined, images of the inside of the shield tunnel are acquired in real time to obtain the first image corresponding to the current moment;

[0042] S2, the first image is subjected to feature extraction using a semantic segmentation algorithm. The encoded features in the first image are extracted, and the encoded features are identified to obtain the digital symbol recognition result corresponding to the encoded features.

[0043] S3, Based on the digital symbol recognition result, determine the location information corresponding to the target electric vehicle at the location to be determined.

[0044] The beneficial effects of the method for determining the location of a battery-powered vehicle in a shield tunnel provided by this invention are as follows:

[0045] First, by using a data acquisition device mounted on an electric vehicle to collect images of the tunnel interior in real time, detailed information about the tunnel's interior can be quickly obtained without being limited by the time and frequency of manual inspection, greatly improving inspection efficiency. Second, semantic segmentation algorithms are used to extract features from the images. Compared to traditional methods, semantic segmentation algorithms can automatically extract higher-level semantic features and more accurately identify coded features in the images, effectively avoiding errors caused by manual identification. Third, the location information of the electric vehicle can be determined based on the digital symbol recognition results, enabling precise positioning of the electric vehicle and facilitating better management and scheduling, thereby improving the efficiency and safety of tunnel construction and maintenance.

[0046] A shield tunnel is a tunnel constructed using the shield tunneling method.

[0047] In addition, in order to improve the accuracy of extracting encoded features from the first image, during the training of the semantic segmentation algorithm, it is necessary to ensure the diversity of the sample set and to rationally set the number of iterations and the loss function.

[0048] Specifically, to ensure the diversity of the sample set, all samples in the set are classified according to the following conditions: no abnormalities, possible occlusion, insufficient lighting, blurred image, and a combination of multiple issues. Occlusion refers to an object obscuring text or symbols. Insufficient lighting refers to an overall sample brightness that is lower than the preset brightness. Blurred image refers to distortion caused by excessively fast movement speed.

[0049] Determine whether the target sample set corresponding to each type of case in the sample set has reached the preset number. If not, the target sample set needs to be expanded.

[0050] The expansion process is as follows:

[0051] When the category that needs to be expanded is occlusion, a target sample image is randomly selected from the corresponding target sample set. The object that occludes the target sample image is randomly rotated or translated. The feature similarity between the first target sample image and the target sample image is continuously calculated after each rotation or translation. When the feature similarity exceeds the preset similarity, the first target sample image is placed in the corresponding target sample set.

[0052] When the category that needs to be expanded is insufficient lighting, a target sample image is randomly selected from the target sample set without abnormalities, and the brightness of the target sample image is adjusted. Each time the brightness is adjusted, the feature recognition ratio between the second target sample image after adjustment and the target sample image is calculated. When the feature recognition ratio is not less than the preset number, the second target sample image is stored in the target sample set corresponding to insufficient lighting.

[0053] When the category requiring expanded processing is a blurred image, a target sample image is randomly selected from the target sample set corresponding to no abnormalities or occlusion. The image is then de-sharpened, and multiple features and corresponding feature boxes are selected from the target sample image. The overlap between the third target sample image obtained after de-sharpening and the features and corresponding feature boxes in the target sample image is determined. When the overlap meets the preset overlap, the third target sample image is placed in the target sample set corresponding to the blurred image.

[0054] When the category that needs to be expanded is a combination of multiple problems, randomly select a target sample image of any category, and randomly perform an expansion scheme for other categories after removing that category once. Try to judge the similarity between the fourth target sample image obtained after the expansion scheme is processed once and the target sample image. If the similarity exceeds the preset similarity, store the fourth target sample image in the corresponding target sample set.

[0055] For example: randomly select any fifth target sample image from the target sample set corresponding to insufficient lighting conditions, add occlusions or reduce the sharpness of the fifth target sample image to generate a fourth target sample image, calculate the similarity between the fourth target sample image and the fifth target sample image, and store the fifth target sample image in the corresponding target sample set when the similarity exceeds the preset similarity.

[0056] Furthermore, the semantic segmentation algorithm performs downsampling on the input data by combining max pooling, average pooling, and convolution operations with a Concat network.

[0057] Furthermore, the process of identifying the encoded features is as follows:

[0058] The encoded features are identified using a digital OCR recognition algorithm;

[0059] The digital OCR recognition algorithm includes: convolutional layers, bidirectional LSTM, and fully connected layers.

[0060] Furthermore, the process of determining the location information corresponding to the target electric vehicle at the location to be determined is as follows:

[0061] By using a pre-defined table of correspondence between target digital symbol recognition results and target location information, the target location information corresponding to the digital symbol recognition results is determined, and the target location information is identified as the location information corresponding to the target electric vehicle at the location to be determined.

[0062] In Example 1, the sensor combination scheme and the data communication between multiple modules in the autonomous driving system are all conducted through the Ethernet protocol, and the time synchronization data acquisition of multiple sensors is realized based on satellite time synchronization.

[0063] Step 1: Annotate the tunnel sidewall image dataset

[0064] First, a large number of coded images of the tunnel sidewalls need to be collected to build a dataset to ensure the accuracy of subsequent training and use.

[0065] The tunnel sidewall coded region and other regions of the dataset are labeled at the pixel level. Since this invention only requires segmentation of the tunnel sidewall coded region, only the numeric regions need to be labeled. The specific labeling range can be found by referring to [reference needed]. Figure 1 Furthermore, to verify the robustness of the tunnel sidewall coding region identification method under different environments, the dataset should include various situations that trains may encounter during operation. Finally, the dataset is divided into training and test sets according to a certain ratio.

[0066] Step 2: Construct a network model for tunnel sidewall coding and recognition

[0067] To achieve high-precision segmentation of the encoded region, this invention designs an encoded region extraction network. The network model consists of two parts: encoding and decoding. Its network layer architecture is as follows: Figure 2 As shown in Table 1, the encoding module of this forward trajectory region recognition model consists of 7 convolutional layers, 4 dilated convolutional layers, and 3 cascaded downsampling operations, while the decoding module consists of 5 deconvolutional layers, 5 convolutional layers, and 1 softmax layer. The specific structure is shown in Table 1.

[0068] Table 1

[0069]

[0070]

[0071] Step 3: Construct cascaded downsampling layers and train the tunnel sidewall coding segmentation model.

[0072] To reduce image information loss during downsampling of tunnel sidewall image features, this invention employs max pooling, average pooling, and convolution operations to downsample the image. To ensure the richness of channel semantic information, a Concat network connection unit is used to fuse the features along the channel dimension. Figure 3 As shown.

[0073] Next, this invention demonstrates model training and testing using PyTorch. The model parameters are trained until the training loss converges, with cross-entropy loss used as the objective function for training the network. The loss is summed across all pixels in a mini-batch. Furthermore, to train the network parameters, this invention uses SGD (stochastic gradient descent) with a fixed learning rate of 0.02 and a momentum of 0.9 to accelerate convergence. Additionally, this invention uses a 10% dropout rate in each convolutional layer of the network.

[0074] Step 4: Construct cascaded downsampling layers and train the tunnel sidewall coding segmentation model.

[0075] Next, this invention designs a digital OCR recognition algorithm to extract encoded numerical values. This invention proposes using the CTPN algorithm, which supports lateral text detection, for digital text recognition. The network structure mainly consists of three parts: convolutional layers, bidirectional LSTM, and fully connected layers, as follows... Figure 4 As shown, a VGG16 base net is used to extract features. The feature map output from conv5 is then scanned using a 3x3 sliding window. This involves creating a 3x3 sliding window on the feature map from conv5, combining each point with its surrounding 3x3 region to obtain a feature vector of length 3x3*C. After reshaping the features, they are input into a bidirectional LSTM layer consisting of two inverse LSTMs. The output of the bidirectional LSTM is then fed into a fully connected layer, introducing an anchor mechanism. For each point, k anchors are used for prediction. Each anchor is a box whose height decreases progressively from [273,…,11], each time divided by 0.7, for a total of 10 anchors.

[0076] Finally, the input enters the fully connected layer. The fully connected layer generates three branches. The first branch outputs 2k vertical coordinates. Since the y-coordinate of an anchor consists of two parts: the center height (y-coordinate) and the height of the bounding box, 2k vertical coordinates are used for each branch. An anchor's y-coordinate has two parts: the center height (y-coordinate) and the height of the bounding box.

[0077]

[0078] in:

[0079] v={v c , v h}and These are the predicted coordinates and the actual coordinates, respectively.

[0080] and h a These are the y-coordinate center and height of an anchor, respectively.

[0081] h and h are the predicted y-coordinate center and height, respectively;

[0082] h* and h* represent the actual y-coordinate center and height, respectively.

[0083] The second branch has 2k scores. Since k text proposals were predicted, there are 2k scores, one for text and one for non-text. (This is a binary classification problem, and the loss function is softmax). If the score is >0.7, the anchor is considered to contain text.

[0084] The third branch consists of k side-refinements, which are mainly used to refine the two endpoints of the text lines, representing the horizontal shift amount for each proposal:

[0085]

[0086] in:

[0087] x side It is the predicted coordinate closest to the anchor's horizontal coordinate (left or right coordinate);

[0088] It is the actual x-coordinate;

[0089] It is the x-coordinate center of the anchor;

[0090] w a That is the width of the anchor, which is 16.

[0091] The RPN network, consisting of two branches, is connected after the fully connected layer to obtain text proposals. For the obtained anchors, the left branch uses bounding box regression to correct the center y-coordinate and height of the text anchors, while the right branch uses softmax to determine whether the anchor contains numbers, i.e., selecting the positive anchor with the highest score. After processing by softmax and bounding box regression in the y-direction, the anchors will produce a set of coded numbers for vertical stripes, such as... Figure 1 The results of the digit recognition are shown below.

[0092] Step 5: Locating the electric vehicle based on tunnel sidewall coding

[0093] After obtaining the tunnel sidewall code, the code number is compared with the pre-determined tunnel sidewall code map to determine the current position of the tunnel sidewall in the entire tunnel excavation, thereby accurately locating the electric vehicle.

[0094] In the above embodiments, although the steps are numbered S1, S2, etc., they are only specific embodiments given by the present invention. Those skilled in the art can adjust the execution order of S1, S2, etc. according to the actual situation, which is also within the protection scope of the present invention. It can be understood that in some embodiments, some or all of the above embodiments may be included.

[0095] This invention also provides a system for determining the location of a battery-powered vehicle inside a shield tunnel, the specific technical solution of which is as follows:

[0096] The acquisition module is used to: acquire images of the inside of the shield tunnel in real time through an acquisition device mounted on a target battery vehicle at the location to be determined, and obtain the first image corresponding to the current moment;

[0097] The recognition module is used to: extract features from the first image using a semantic segmentation algorithm, extract coded features from the first image, recognize the coded features, and obtain the recognition result of the digital symbol corresponding to the coded features;

[0098] The determining module is used to: determine the location information corresponding to the target electric vehicle at the location to be determined based on the digital symbol recognition result.

[0099] Based on the above solution, the present invention can be further improved as follows.

[0100] Furthermore, the semantic segmentation algorithm performs downsampling on the input data by combining max pooling, average pooling, and convolution operations with a Concat network.

[0101] Furthermore, the process of identifying the encoded features is as follows:

[0102] The encoded features are identified using a digital OCR recognition algorithm;

[0103] The digital OCR recognition algorithm includes: convolutional layers, bidirectional LSTM, and fully connected layers.

[0104] Furthermore, the process of determining the location information corresponding to the target electric vehicle at the location to be determined is as follows:

[0105] By using a pre-defined table of correspondence between target digital symbol recognition results and target location information, the target location information corresponding to the digital symbol recognition results is determined, and the target location information is identified as the location information corresponding to the target electric vehicle at the location to be determined.

[0106] It should be noted that the beneficial effects of the system for determining the position of a battery-powered vehicle in a shield tunnel provided in the above embodiments are the same as those of the method for determining the position of a battery-powered vehicle in a shield tunnel described above, and will not be repeated here. Furthermore, the system provided in the above embodiments is only illustrated by the division of the above functional modules. In practical applications, the above functions can be assigned to different functional modules as needed, that is, the system can be divided into different functional modules according to the actual situation to complete all or part of the functions described above. In addition, the system and method embodiments provided in the above embodiments belong to the same concept, and their specific implementation process is detailed in the method embodiments, and will not be repeated here.

[0107] like Figure 5 As shown, an electronic device 300 according to an embodiment of the present invention includes a processor 320 coupled to a memory 310. The memory 310 stores at least one computer program 330, which is loaded and executed by the processor 320 to enable the electronic device 300 to implement any of the above-mentioned methods. Specifically:

[0108] The electronic device 300 can vary considerably due to differences in configuration or performance. It may include one or more processors 320 (Central Processing Units, CPUs) and one or more memories 310. The memories 310 store at least one computer program 330, which is loaded and executed by the processors 320 to enable the electronic device 300 to implement the method for determining the position of a battery-powered vehicle within a shield tunnel as described in the above embodiment. Of course, the electronic device 300 may also have wired or wireless network interfaces, a keyboard, and input / output interfaces for input and output. It may also include other components for implementing device functions, which will not be elaborated upon here.

[0109] An embodiment of the present invention provides a computer-readable storage medium storing at least one computer program, which is loaded and executed by a processor to enable a computer to implement any of the above-described methods.

[0110] Alternatively, the computer-readable storage medium may be a read-only memory (ROM), a random access memory (RAM), a compact disc read-only memory (CD-ROM), magnetic tape, a floppy disk, and an optical data storage device, etc.

[0111] In an exemplary embodiment, a computer program product or computer program is also provided, which includes computer instructions stored in a computer-readable storage medium. A processor of an electronic device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the electronic device to perform any of the methods described above.

[0112] It should be noted that the terms "first," "second," etc., used in the specification and claims of this application are used to distinguish similar objects and represent a limitation on a specific order or sequence. Where appropriate, the order of use for similar objects can be interchanged so that the embodiments of this application described herein can be implemented in an order other than that shown or described.

[0113] Those skilled in the art will recognize that this invention can be implemented as a system, method, or computer program product. Therefore, this disclosure can be specifically implemented in the following forms: it can be entirely hardware, entirely software (including firmware, resident software, microcode, etc.), or a combination of hardware and software, generally referred to herein as a "circuit," "module," or "system." Furthermore, in some embodiments, this invention can also be implemented as a computer program product in one or more computer-readable media containing computer-readable program code.

[0114] Any combination of one or more computer-readable media may be used. A computer-readable medium can be a computer-readable signal medium or a computer-readable storage medium. A computer-readable storage medium can be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples (a non-exhaustive list) of computer-readable storage media include: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In this document, a computer-readable storage medium can be any tangible medium that contains or stores a program that can be used by or in connection with an instruction execution system, apparatus, or device.

[0115] Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of the present invention.

Claims

1. A method for determining the location of a battery-powered vehicle inside a shield tunnel, characterized in that, include: By using the acquisition device mounted on the target battery vehicle at the location to be determined, images of the inside of the shield tunnel are acquired in real time to obtain the first image corresponding to the current moment. The first image is subjected to feature extraction using a semantic segmentation algorithm. The encoded features in the first image are extracted, and the encoded features are identified to obtain the digital symbol recognition result corresponding to the encoded features. Based on the digital symbol recognition results, the location information corresponding to the target electric vehicle at the location to be determined is determined.

2. The method for determining the location of a battery-powered vehicle in a shield tunnel according to claim 1, characterized in that, The semantic segmentation algorithm downsamples the input data by combining max pooling, average pooling, and convolution operations with a Concat network.

3. The method for determining the location of a battery-powered vehicle in a shield tunnel according to claim 1, characterized in that, The process of identifying the encoded features is as follows: The encoded features are identified using a digital OCR recognition algorithm; The digital OCR recognition algorithm includes: convolutional layers, bidirectional LSTM, and fully connected layers.

4. The method for determining the location of a battery-powered vehicle in a shield tunnel according to claim 1, characterized in that, The process of determining the location information corresponding to the target electric vehicle at the location to be determined is as follows: By using a pre-defined table of correspondence between target digital symbol recognition results and target location information, the target location information corresponding to the digital symbol recognition results is determined, and the target location information is identified as the location information corresponding to the target electric vehicle at the location to be determined.

5. A system for determining the location of a battery-powered vehicle inside a shield tunnel, characterized in that, include: The acquisition module is used to: acquire images of the inside of the shield tunnel in real time through an acquisition device mounted on a target battery vehicle at the location to be determined, and obtain the first image corresponding to the current moment; The recognition module is used to: extract features from the first image using a semantic segmentation algorithm, extract coded features from the first image, recognize the coded features, and obtain the recognition result of the digital symbol corresponding to the coded features; The determining module is used to: determine the location information corresponding to the target electric vehicle at the location to be determined based on the digital symbol recognition result.

6. The system for determining the location of a battery-powered vehicle in a shield tunnel according to claim 5, characterized in that, The semantic segmentation algorithm downsamples the input data by combining max pooling, average pooling, and convolution operations with a Concat network.

7. The system for determining the location of a battery-powered vehicle in a shield tunnel according to claim 5, characterized in that, The process of identifying the encoded features is as follows: The encoded features are identified using a digital OCR recognition algorithm; The digital OCR recognition algorithm includes: convolutional layers, bidirectional LSTM, and fully connected layers.

8. The system for determining the location of a battery-powered vehicle in a shield tunnel according to claim 5, characterized in that, The process of determining the location information corresponding to the target electric vehicle at the location to be determined is as follows: By using a pre-defined table of correspondence between target digital symbol recognition results and target location information, the target location information corresponding to the digital symbol recognition results is determined, and the target location information is identified as the location information corresponding to the target electric vehicle at the location to be determined.

9. An electronic device, characterized in that, The electronic device includes a processor coupled to a memory storing at least one computer program, which is loaded and executed by the processor to enable the electronic device to perform the method as described in any one of claims 1 to 4.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores at least one computer program, which is loaded and executed by a processor to enable the computer to perform the method as described in any one of claims 1 to 4.

Images