Polygon centerline extraction method, device, and guide point determination method

Abstract

The present disclosure relates to the fields of data processing technology, specifically to map navigation, artificial intelligence, and other technical fields, and more particularly to methods, devices, and guide point determination methods for extracting polygon centerlines. A specific implementation involves obtaining a polygon's contour point sequence; inputting this contour point sequence into a centerline extraction model trained using a neural network; the centerline extraction model outputs a polygon's centerline point sequence, and based on this centerline point sequence, the polygon's centerline is obtained. The centerline extraction model, trained through deep learning, can predict the polygon's centerline based on its contour points, resulting in an accurate centerline prediction using a simple algorithm.

Description

Technical Field

[0001] The present disclosure relates to the field of data processing technology, specifically to technical fields such as map navigation and artificial intelligence, and in particular to a method and device for extracting polygon centerlines and a method for determining guide points. Background Art

[0002] Driving is a very common and important mode of transportation. How to accurately and efficiently guide drivers to the target POI (Point of Interest) is one of the core competitive advantages of map navigation products. There are rivers, lakes and other water bodies in the map. Guidance points that cross rivers and lakes will lead to a poor user experience. Therefore, it is necessary to filter the guidance points that cross rivers and lakes on the line based on the river and lake data. Directly using the raw data of rivers and lakes has the following problems: (1) River data is confidential data for maps, and the full data cannot be used directly; (2) Some river and lake data contain a large number of shape points, which makes the algorithm inefficient.

[0003] Currently, the method of extracting the center lines of rivers and lakes is used to filter high-risk guidance points. However, the existing center line solutions cannot take into account both efficiency and accuracy. The first is the Voronoi diagram solution. Although the center lines produced are accurate, the algorithm complexity is high and a large number of similar shape points will be produced. The second is the solution of finding the midpoint of two intersection points of parallel lines. Although this solution is efficient, it performs poorly in finding shape points of concave polygons and has low accuracy. Summary of the Invention

[0004] The present disclosure provides a polygon centerline extraction method, device, guide point determination method, electronic device, storage medium, and computer program product.

[0005] According to a first aspect of the present disclosure, a method for extracting a polygon centerline is provided, comprising:

[0006] Get the polygon's contour point sequence;

[0007] The contour point sequence is input into a centerline extraction model obtained based on neural network training. The centerline extraction model outputs the centerline point sequence of the polygon, and the centerline of the polygon is obtained based on the centerline point sequence.

[0008] According to a second aspect of the present disclosure, a model training method is provided, comprising:

[0009] Get the polygon contour point sequence and target centerline point sequence;

[0010] The contour point sequence is used as the input of a neural network, the target centerline point sequence is used as the output of the neural network, and the neural network is trained to obtain a centerline extraction model; the centerline extraction model is used to extract the centerline point sequence of the polygon.

[0011] According to a third aspect of the present disclosure, a guidance point determination method is provided, comprising:

[0012] Get the target point of interest;

[0013] Determining a navigation track point based on the target point of interest;

[0014] The navigation track points are filtered based on the center line obtained by the extraction method described in any one of the above technical solutions to obtain guide points.

[0015] According to a fourth aspect of the present disclosure, there is provided a device for extracting a polygon centerline, comprising:

[0016] A first acquisition module is configured to acquire a sequence of contour points of a polygon;

[0017] The centerline extraction module is configured to input the contour point sequence into a centerline extraction model obtained based on neural network training, output the centerline point sequence of the polygon through the centerline extraction model, and obtain the centerline of the polygon based on the centerline point sequence.

[0018] According to a fifth aspect of the present disclosure, a model training device is provided, comprising:

[0019] A second acquisition module is configured to acquire a polygon contour point sequence and a target centerline point sequence;

[0020] The model training module is configured to use the contour point sequence as the input of the neural network and the target centerline point sequence as the output of the neural network, and train the neural network to obtain a centerline extraction model; the centerline extraction model is used to extract the centerline point sequence of the polygon.

[0021] According to a sixth aspect of the present disclosure, there is provided a guidance point determination device, comprising:

[0022] a third acquisition module, configured to acquire a target point of interest;

[0023] a determination module, configured to determine a navigation track point based on the target point of interest;

[0024] The filtering module is configured to filter the navigation trajectory points based on the center line obtained by the method according to any one of claims 1 to 6 to obtain guidance points.

[0025] According to the seventh aspect of the present disclosure, an electronic device is provided, comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute any one of the methods described in the above technical solutions.

[0026] According to an eighth aspect of the present disclosure, a non-transitory computer-readable storage medium storing computer instructions is provided, wherein the computer instructions are used to enable the computer to execute the method according to any one of the above technical solutions.

[0027] According to a ninth aspect of the present disclosure, a computer program product is provided, comprising a computer program, which implements the method according to any one of the above technical solutions when executed by a processor.

[0028] The present disclosure provides a method and device for extracting the centerline of a polygon, and a method for determining a guide point. The centerline extraction model obtained through deep learning training can predict the centerline of a polygon based on its contour points. The produced centerline is accurate and the algorithm is simple. The method can be used to determine guide points during navigation, filter high-risk guide points, and enhance the user experience.

[0029] It should be understood that the contents described in this section are not intended to identify the key or important features of the embodiments of the present disclosure, nor are they intended to limit the scope of the present disclosure. Other features of the present disclosure will become readily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] The accompanying drawings are provided to facilitate a better understanding of the present invention and do not constitute a limitation of the present disclosure.

[0031] Figure 1 1 is a schematic diagram of the steps of a method for extracting a polygon centerline in an embodiment of the present disclosure;

[0032] Figure 2 Schematic diagram of the principle of the method for extracting the center line of a polygon in an embodiment of the present disclosure;

[0033] Figure 3 is a schematic diagram of the steps of the model training method in an embodiment of the present disclosure;

[0034] Figure 4 is a structural diagram of a centerline extraction model in an embodiment of the present disclosure;

[0035] Figure 5 is a schematic diagram of the steps of the guidance point determination method in an embodiment of the present disclosure;

[0036] Figure 6is a schematic diagram of the principle of the guidance point determination method in an embodiment of the present disclosure;

[0037] Figure 7 is a principle block diagram of a device for extracting a polygon centerline in an embodiment of the present disclosure;

[0038] Figure 8 is a principle block diagram of the model training device in an embodiment of the present disclosure;

[0039] Figure 9 is a principle block diagram of a guidance point determination device in an embodiment of the present disclosure;

[0040] Figure 10 is a schematic block diagram of an example electronic device in an embodiment of the present disclosure. DETAILED DESCRIPTION

[0041] The following description of exemplary embodiments of the present disclosure is made in conjunction with the accompanying drawings, including various details of the embodiments of the present disclosure to facilitate understanding. These details should be considered as merely exemplary. Therefore, those skilled in the art will recognize that various changes and modifications may be made to the embodiments described herein without departing from the scope and spirit of the present disclosure. Similarly, for the sake of clarity and conciseness, descriptions of well-known functions and structures are omitted in the following description.

[0042] In order to solve the technical problem that the centerline extraction scheme in the prior art cannot take into account both efficiency and accuracy, the present invention provides a method for extracting the centerline of a polygon, such as Figure 1 Shown, including:

[0043] Step S101, obtain the polygon contour point sequence. Figure 2 As shown, a polygon's contour point sequence consists of multiple contour points. Contour points are typically manually labeled and stored in a map database. Contour point coordinates can be directly retrieved from the map database: (x0, y0) -> (x1, y1) -> ... -> (xn, yn) -> (x0, y0). For a river or lake, the polygon extracted is a closed shape. The first and last points in the contour point sequence are the same point, (x0, y0). The contour point sequence can be selected from representative contour points of the polygon, sufficient to describe the shape of the river or lake.

[0044] In step S102, the contour point sequence is input into a centerline extraction model trained using a neural network. The centerline extraction model outputs a polygon centerline point sequence and then determines the polygon's centerline based on this centerline point sequence. The centerline extraction model is a deep learning model trained using a neural network. It can directly predict the centerline point sequence based on the input contour point sequence, offering a simple algorithm and high prediction accuracy.

[0045] As an optional implementation, in step S102, the contour point sequence is input into a centerline extraction model obtained through neural network training. The centerline extraction model outputs a centerline point sequence of the polygon including:

[0046] Input the contour point sequence into the centerline extraction model.

[0047] The neural network in this embodiment can be a seq2seq (sequence to sequence) model, and the structure diagram is as follows: Figure 4 As shown, it can be an encoder-decoder structure or an LSTM (Long Short-Term Memory) network.

[0048] The centerline extraction model encodes the contour point sequence to obtain the corresponding encoding information.

[0049] For example, if the centerline extraction model uses an encoder-decoder structure, the encoder module can be used to encode the contour point sequence to obtain encoding information.

[0050] The centerline extraction model performs decoding processing based on the encoded information to obtain corresponding decoded information.

[0051] For example, if the centerline extraction model uses an encoder-decoder structure, the decoder module can be used to decode the encoded information to obtain the decoded information.

[0052] Generate the corresponding centerline point sequence according to the decoded information.

[0053] As an optional implementation, before step S102 inputs the contour point sequence into the centerline extraction model, it also includes: normalizing the contour point sequence and inputting the normalized contour point sequence into the centerline extraction model. After step S102 outputs the centerline point sequence of the polygon through the centerline extraction model, and before obtaining the centerline of the polygon based on the centerline point sequence, it also includes: inverse normalization of the centerline point sequence. Normalization is to limit the data to be processed to a certain range after processing. First, normalization is for the convenience of subsequent data processing, and secondly, to ensure faster convergence when the program is running. The specific role of normalization is to summarize the statistical distribution of unified samples. Normalization between 0-1 is a statistical probability distribution, and normalization in a certain interval is a statistical coordinate distribution. Accordingly, after the model outputs the centerline point sequence, the centerline point sequence must also be inverse normalized to restore the centerline points.

[0054] As an optional implementation, normalizing the contour point sequence includes:

[0055] Determine the maximum and minimum values ​​of the coordinates of multiple contour points in a contour point sequence. Find the minimum and maximum values ​​of x and y, respectively, for all contour point coordinates: x_min = min(x0,x1,…,xn), y_max = max(y0,y1,…,yn), x_max = max(x0,x1,…,xn), y_min = min(y0,y1,…,yn).

[0056] Normalize each contour point's coordinates based on their maximum and minimum values, limiting them to a preset numerical range. [xi_scaled,yi_scaled] = [xi*(xi-x_min) / (x_max-x_min)),yi*(yi-y_min) / (y_max-y_min)]. The preset numerical range can be between 0 and 1.

[0057] As an optional implementation, performing inverse normalization processing on the centerline point sequence includes:

[0058] Determine the maximum and minimum values ​​of the coordinates of multiple contour points in a contour point sequence, i.e. x_min, x_max, y_min, y_max.

[0059] Perform a reverse normalization calculation on each centerline point in the centerline point sequence based on the maximum and minimum values ​​of the contour point coordinates to restore each centerline point coordinate. The reverse normalization calculation is performed on each centerline point using the following formula: [Txi_recover, Tyi_recover] = [x_min + Txi*(x_max-x_min)), y_min + Tyi*(y_max-y_min)].

[0060] The present disclosure also provides a model training method, such as Figure 3 Shown, including:

[0061] Step S301, obtain the polygon contour point sequence and the target center line point sequence. Figure 2As shown, the polygon's contour point coordinates (x0, y0) -> (x1, y1) -> ... -> (xn, yn) -> (x0, y0) and the manually labeled centerline point coordinates (Tx0, Ty0), (Tx1, Ty1) ... (Txm, Tym) are obtained as training data. The target centerline point sequence serves as the target parameter for training the neural network. Before obtaining the centerline point sequence, the centerline points can be manually labeled. The centerline point sequence is kept as sparse as possible, retaining only representative centerline points. For example, if there are 10 million centerline points, 10,000 of them can be selected, or one thousandth of the centerline points. This maximizes model learning efficiency while ensuring model calculation accuracy.

[0062] Step S302: Take the contour point sequence as the input of the neural network and the target centerline point sequence as the output of the neural network, and train the neural network to obtain a centerline extraction model; the centerline extraction model is used to extract the centerline point sequence of the polygon. The neural network in this embodiment can be a seq2seq model, and the structure diagram is as follows: Figure 4 As shown in Figure 2, the sequence-to-sequence model C can be either an LSTM or an encoder-decoder structure. The contour point sequence is used as the input sample of the seq2seq model, which predicts the centerline point sequence based on the contour point sequence until the predicted centerline point sequence is close to the target centerline point sequence.

[0063] As an optional implementation, step S302, using the contour point sequence as the input of the neural network and the target centerline point sequence as the output of the neural network, and training the neural network to obtain a centerline extraction model includes:

[0064] Input the contour point sequence into the neural network. For example, the neural network in this embodiment can be a seq2seq model, and the structure diagram is as follows: Figure 4 As shown, the seq2seq model C can be an encoder-decoder structure or an LSTM.

[0065] The neural network encodes the contour point sequence to obtain corresponding encoding information. For example, if the neural network uses an encoder-decoder structure, the encoder module can be used to encode the contour point sequence to obtain encoding information.

[0066] The neural network decodes the encoded information to obtain the corresponding decoded information. If the neural network uses an encoder-decoder structure, the decoder module can be used to decode the encoded information to obtain the decoded information.

[0067] The difference between the decoded information and the target centerline point sequence is calculated until the difference reaches the target value, completing the training of the neural network and obtaining the centerline extraction model.

[0068] In this embodiment, the target centerline point sequence is the training target of the model. The encoding phase can be understood as the encoder module converting the input sequence into a fixed-length vector, and the decoding phase can be understood as the decoder module decoding the input fixed-length vector into an output sequence. Figure 4 When the sequence-to-sequence model c shown is trained, the encoder module can generate a state vector (i.e., encoding information) of the sample contour point sequence, and the decoder module can generate a centerline point sequence (i.e., decoding information) based on the state vector. During the model training process, the parameters of the neural network can be adjusted based on a preset real centerline point sequence (i.e., target centerline point sequence). After a certain number of training iterations, the centerline point sequence predicted by the neural network based on the contour point sequence gradually approaches the target centerline point sequence, and the neural network converges to obtain a centerline extraction model.

[0069] As an optional implementation, in step S302, the contour point sequence is used as the input of the neural network, and the target centerline point sequence is used as the output of the neural network. Before the neural network is trained to obtain the centerline extraction model, the contour point sequence and the target centerline point sequence can be normalized. Normalization is performed to facilitate subsequent data processing and to ensure faster model convergence.

[0070] The present disclosure provides a method for determining a guidance point. Figure 5 Shown, including:

[0071] Step S501: Obtain a target point of interest. The target point of interest is related to the user's target navigation destination. The point of interest can be a shopping mall, hospital, bus station, etc.

[0072] Step S502: determining a navigation track point based on the target point of interest. A guide point belongs to the navigation track point and refers to a navigation guidance destination based on the target point of interest.

[0073] Step S503, based on the center line of the lake or river extracted by the polygon center line extraction method described in any of the above embodiments, the navigation trajectory points are filtered through the center line to obtain the guide point. If during the navigation process, the guide point crosses the lake or river, it will cause inconvenience to the user (the user cannot drive directly across the lake or river and can only take a longer detour to reach the guide point). Therefore, in this embodiment, the center line of the lake or river is extracted, and the high-risk guide points are filtered by judging whether there is an intersection between the center line of the lake or river and the navigation trajectory formed by the navigation trajectory point, thereby screening out the required guide points and avoiding the guide points crossing the lake or river, which brings a bad experience to the user. Figure 6 As shown, assuming that the user's target point of interest is point A, the system determines several navigation track points B and C based on the target point of interest being point A. The user's current location O and the navigation track points form a navigation track. There is a river R in the map. If there is an intersection between the navigation track and the river R, it means that the navigation track point is a high-risk guidance point across the river. For example, if the navigation track between O and C has an intersection X with the river R, then C is a high-risk guidance point and the system filters it.

[0074] The present disclosure also provides a polygon centerline extraction device 700, such as Figure 7 Shown, including:

[0075] The first acquisition module 701 is configured to acquire a sequence of polygonal contour points. Figure 2 As shown, a polygon's contour point sequence consists of multiple contour points. Contour points are typically manually labeled and stored in a map database. Contour point coordinates can be directly retrieved from the map database: (x0, y0) -> (x1, y1) -> ... -> (xn, yn) -> (x0, y0). For a river or lake, the polygon extracted is a closed shape. The first and last points in the contour point sequence are the same point, (x0, y0). The contour point sequence can be selected from representative contour points of the polygon, sufficient to describe the shape of the river or lake.

[0076] Centerline extraction module 702 is configured to input a contour point sequence into a centerline extraction model trained using a neural network. The centerline extraction model outputs a polygon centerline point sequence, and then obtains the polygon's centerline based on the centerline point sequence. The centerline extraction model, trained using a neural network, can directly predict the centerline point sequence based on the input contour point sequence. This model has a simple algorithm and high prediction accuracy.

[0077] As an optional implementation, the centerline extraction module 702 inputs the contour point sequence into a centerline extraction model obtained through neural network training. The centerline extraction model outputs a centerline point sequence of the polygon including:

[0078] Input the contour point sequence into the centerline extraction model.

[0079] The neural network in this embodiment can be a seq2seq model, the structure diagram is as follows Figure 4 As shown, it can be an encoder-decoder structure or an LSTM.

[0080] The centerline extraction model encodes the contour point sequence to obtain the corresponding encoding information.

[0081] For example, if the centerline extraction model uses an encoder-decoder structure, the encoder module can be used to encode the contour point sequence to obtain encoding information.

[0082] The centerline extraction model performs decoding processing based on the encoded information to obtain corresponding decoded information.

[0083] For example, if the centerline extraction model uses an encoder-decoder structure, the decoder module can be used to decode the encoded information to obtain the decoded information.

[0084] Generate the corresponding centerline point sequence according to the decoded information.

[0085] As an optional embodiment, the extraction device further includes:

[0086] The first processing module is configured to perform normalization processing on the contour point sequence before the contour point sequence is input into the centerline extraction model.

[0087] The second processing module is configured as a centerline extraction module to perform inverse normalization processing on the centerline point sequence before obtaining the centerline of the polygon based on the centerline point sequence.

[0088] Specifically, normalization involves limiting the processed data to a certain range. Normalization facilitates subsequent data processing and ensures faster convergence during program execution. The specific function of normalization is to summarize and unify the statistical distribution of samples. Normalization between 0 and 1 represents a statistical probability distribution, while normalization within a certain interval represents a statistical coordinate distribution. Accordingly, after the model outputs a sequence of centerline points and before deriving the centerline of a polygon based on this sequence, the sequence must be denormalized to restore the centerline points.

[0089] As an optional implementation, the first processing module includes:

[0090] The first determining unit is configured to determine the maximum and minimum values ​​of the coordinates of multiple contour points in the contour point sequence. The minimum and maximum values ​​of x and y in the coordinates of all contour points are found, respectively, x_min = min(x0, x1, ..., xn), y_max = max(y0, y1, ..., yn), x_max = max(x0, x1, ..., xn), and y_min = min(y0, y1, ..., yn).

[0091] The first calculation unit is configured to perform normalization calculation on each contour point coordinate based on the maximum and minimum values ​​of the contour point coordinates, and limit each contour point coordinate to a preset numerical range. [xi_scaled,yi_scaled] = [xi*(xi-x_min) / (x_max-x_min)),yi*(yi-y_min) / (y_max-y_min)]. The preset numerical range can be between 0 and 1.

[0092] As an optional implementation, the second processing module includes:

[0093] The second determining unit is configured to determine the maximum value and the minimum value of the coordinates of multiple contour points in the contour point sequence, namely x_min, x_max, y_min, y_max.

[0094] The second calculation unit is configured to perform a reverse normalization calculation on the coordinates of each centerline point in the centerline point sequence based on the maximum and minimum values ​​of the contour point coordinates, and restore the coordinates of each centerline point. The reverse normalization calculation is performed on each centerline point using the following formula: [Txi_recover, Tyi_recover] = [x_min + Txi * (x_max - x_min)), y_min + Tyi * (y_max - y_min)].

[0095] The present disclosure also provides a model training device 800, such as Figure 8 Shown, including:

[0096] The second acquisition module 801 is configured to acquire the polygon contour point sequence and the target center line point sequence. Figure 2As shown, the coordinates of the polygon's contour points and the manually marked centerline points are obtained as training data. The target centerline point sequence serves as the target parameter in the neural network training process. Before obtaining the centerline point sequence, the centerline points can be obtained through manual marking. The centerline point sequence is as sparse as possible, retaining only representative centerline points. For example, if there are 10 million centerline points, 10,000 of them can be selected, that is, 1 / 1000 of the centerline points. This maximizes the model's learning efficiency while ensuring the accuracy of the model calculation.

[0097] The model training module 802 is configured to use the contour point sequence as the input of the neural network and the target centerline point sequence as the output of the neural network, and train the neural network to obtain a centerline extraction model; the centerline extraction model is used to extract the centerline point sequence of the polygon. The neural network in this embodiment can be a seq2seq model, and the structure diagram is as follows: Figure 4 As shown in Figure 1, it can be composed of an encoder module and a decoder module, or it can be an LSTM. The contour point sequence is used as the input sample of the seq2seq model, and the seq2seq model predicts the centerline point sequence based on the contour point sequence until the predicted centerline point sequence is close to the target centerline point sequence.

[0098] As an optional implementation, the model training module 802 includes:

[0099] The input unit is configured to input the contour point sequence into the neural network. For example, the neural network in this embodiment can be a seq2seq model, and the structure diagram is as follows: Figure 4 As shown, the seq2seq model C can be an encoder-decoder structure or an LSTM.

[0100] The encoding unit is configured to encode the contour point sequence using a neural network to obtain corresponding encoding information. For example, if the neural network uses an encoder-decoder structure, the encoder module can be used to encode the contour point sequence to obtain the encoding information.

[0101] The decoding unit is configured to decode the encoded information using a neural network to obtain corresponding decoded information. If the neural network uses an encoder-decoder structure, the decoder module can be used to decode the encoded information to obtain the decoded information.

[0102] The computing unit is configured to calculate the difference between the decoded information and the target centerline point sequence until the difference reaches the target value, thereby completing the training of the neural network and obtaining a centerline extraction model.

[0103] In this embodiment, the target centerline point sequence is the training target of the model. The encoding phase can be understood as the encoder module converting the input sequence into a fixed-length vector, and the decoding phase can be understood as the decoder module decoding the input fixed-length vector into an output sequence. Figure 4 When the sequence-to-sequence model c shown is trained, the encoder module can generate a state vector (i.e., encoding information) of the sample contour point sequence, and the decoder module can generate a centerline point sequence (i.e., decoding information) based on the state vector. During the model training process, the parameters of the neural network can be adjusted based on a preset real centerline point sequence (i.e., target centerline point sequence). After a certain number of training iterations, the centerline point sequence predicted by the neural network based on the contour point sequence gradually approaches the target centerline point sequence, and the neural network converges to obtain a centerline extraction model.

[0104] As an optional embodiment, the model training apparatus further includes: a third processing module configured as a model training module 802. The model training module 802 uses the contour point sequence as input to the neural network and the target centerline point sequence as output of the neural network. Before training the neural network to obtain the centerline extraction model, the contour point sequence and the target centerline point sequence may be normalized. Normalization is performed to facilitate subsequent data processing and, secondly, to ensure accelerated model convergence.

[0105] The present disclosure also provides a guidance point determination device 900, such as Figure 9 Shown, including:

[0106] The third acquisition module 901 is configured to acquire a target point of interest. The target point of interest is related to the user's target navigation destination. The point of interest can be a shopping mall, hospital, bus station, etc.

[0107] The determination module 902 is configured to determine a navigation track point based on the target point of interest. A guidance point belongs to the navigation track point and refers to a navigation guidance end point based on the target point of interest.

[0108] Filtering module 903 is configured to filter navigation trajectory points using the centerline extracted from a lake or river using the polygon centerline extraction method described in any of the above embodiments to obtain a guide point. The filtering method used in the filtering module is the same as the guide point determination method and will not be further described below.

[0109] In the technical solutions disclosed herein, the acquisition, storage, and application of user personal information involved comply with the provisions of relevant laws and regulations and do not violate public order and good morals.

[0110] According to an embodiment of the present disclosure, the present disclosure also provides an electronic device, a readable storage medium, and a computer program product.

[0111] Figure 10 A schematic block diagram of an example electronic device 1000 that can be used to implement embodiments of the present disclosure is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital assistants, cellular phones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present disclosure described and / or claimed herein.

[0112] like Figure 10 As shown, the device 1000 includes a computing unit 1001, which can perform various appropriate actions and processes according to a computer program stored in a read-only memory (ROM) 1002 or a computer program loaded from a storage unit 1008 into a random access memory (RAM) 1003. Various programs and data required for the operation of the device 1000 can also be stored in the RAM 1003. The computing unit 1001, the ROM 1002, and the RAM 1003 are connected to each other via a bus 1004. An input / output (I / O) interface 1005 is also connected to the bus 1004.

[0113] Various components in device 1000 are connected to I / O interface 1005, including an input unit 1006, such as a keyboard, mouse, etc.; an output unit 1007, such as various types of displays, speakers, etc.; a storage unit 1008, such as a magnetic disk, optical disk, etc.; and a communication unit 1009, such as a network card, modem, wireless communication transceiver, etc. The communication unit 1009 allows device 1000 to exchange information / data with other devices via a computer network such as the Internet and / or various telecommunication networks.

[0114] The computing unit 1001 can be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the computing unit 1001 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various computing units that run machine learning objective function algorithms, digital signal processors (DSPs), and any appropriate processors, controllers, microcontrollers, etc. The computing unit 1001 performs the various methods and processes described above, such as the method. For example, in some embodiments, the method can be implemented as a computer software program that is tangibly contained in a machine-readable medium, such as the storage unit 1008. In some embodiments, part or all of the computer program can be loaded and / or installed on the device 1000 via the ROM 1002 and / or the communication unit 1009. When the computer program is loaded into the RAM 1003 and executed by the computing unit 1001, one or more steps of the method described above can be performed. Alternatively, in other embodiments, the computing unit 1001 can be configured to perform the method in any other appropriate manner (e.g., by means of firmware).

[0115] Various embodiments of the systems and techniques described herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), system-on-chip systems (SOCs), programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments can include being implemented in one or more computer programs that are executable and / or interpreted on a programmable system comprising at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0116] The program code for implementing the method of the present disclosure can be written in any combination of one or more programming languages. These program codes can be provided to a processor or controller of a general-purpose computer, a special-purpose computer, or other programmable data processing device so that when the program code is executed by the processor or controller, the functions / operations specified in the flow chart and / or block diagram are implemented. The program code can be executed entirely on the machine, partially on the machine, as a stand-alone software package, partially on the machine and partially on a remote machine, or entirely on a remote machine or server.

[0117] In the context of the present disclosure, a machine-readable medium can be a tangible medium that can contain or store a program for use by or in conjunction with an instruction execution system, device or equipment. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium can include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or equipment, or any suitable combination of the foregoing. A more specific example of a machine-readable storage medium can include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0118] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user can provide input to the computer. Other types of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, voice input, or tactile input).

[0119] The systems and techniques described herein can be implemented in a computing system that includes back-end components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes front-end components (e.g., a user computer having a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such back-end components, middleware components, or front-end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include a local area network (LAN), a wide area network (WAN), and the Internet.

[0120] A computer system may include a client and a server. The client and server are generally remote from each other and typically interact through a communication network. The client-server relationship arises through computer programs running on the respective computers and having a client-server relationship with each other. The server may be a cloud server, a server in a distributed system, or a server integrated with a blockchain.

[0121] It should be understood that the various forms of the processes shown above can be used to reorder, add, or delete steps. For example, the steps described in this disclosure can be performed in parallel, sequentially, or in a different order, as long as the desired results of the technical solutions disclosed in this disclosure can be achieved. This is not limited herein.

[0122] The above specific embodiments do not constitute a limitation on the scope of protection of this disclosure. Those skilled in the art will appreciate that various modifications, combinations, sub-combinations, and substitutions may be made based on design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this disclosure shall be included within the scope of protection of this disclosure.

Claims

1. A method for extracting the centerline of a polygon, wherein: The method is applied in the field of map navigation, including: Obtaining a polygonal contour point sequence; wherein the polygonal contour point sequence is used to describe the shape of a river or a lake; The contour point sequence is input into a centerline extraction model obtained based on neural network training. The centerline extraction model outputs the centerline point sequence of the polygon, and the centerline of the polygon is obtained based on the centerline point sequence.

2. The method according to claim 1, before inputting the contour point sequence into the centerline extraction model, further comprising: performing normalization processing on the contour point sequence; Before obtaining the center line of the polygon based on the center line point sequence, the method further includes: Perform inverse normalization processing on the centerline point sequence.

3. The method according to claim 2, wherein: The normalizing process for the contour point sequence comprises: Determining a maximum value and a minimum value among coordinates of a plurality of contour points in the contour point sequence; The coordinates of each contour point are normalized based on the maximum value and the minimum value of the contour point coordinates, so that the coordinates of each contour point are limited to a preset value range.

4. The method according to claim 2, wherein: The performing inverse normalization processing on the centerline point sequence includes: Determining a maximum value and a minimum value among coordinates of a plurality of contour points in the contour point sequence; Based on the maximum value and the minimum value of the contour point coordinates, an inverse normalization calculation is performed on the coordinates of each center line point in the center line point sequence to restore the coordinates of each center line point.

5. The method according to any one of claims 1 to 4, wherein: Inputting the contour point sequence into a centerline extraction model obtained through neural network training, and the centerline extraction model outputting the centerline point sequence of the polygon comprises: Inputting the contour point sequence into the centerline extraction model; The centerline extraction model performs encoding processing on the contour point sequence to obtain corresponding encoding information; The centerline extraction model performs decoding processing on the encoded information to obtain corresponding decoded information; The corresponding centerline point sequence is generated according to the decoding information.

6. A model training method, wherein: The method is applied in the field of map navigation, including: Obtaining a polygonal contour point sequence and a target centerline point sequence; wherein the polygonal contour point sequence is used to describe the shape of a river or a lake; The contour point sequence is used as the input of a neural network, the target centerline point sequence is used as the output of the neural network, and the neural network is trained to obtain a centerline extraction model; the centerline extraction model is used to extract the centerline point sequence of the polygon.

7. The method according to claim 6, wherein: The step of using the contour point sequence as an input of a neural network, using the target centerline point sequence as an output of the neural network, and training the neural network to obtain a centerline extraction model comprises: Inputting the contour point sequence into the neural network; The neural network performs encoding processing on the contour point sequence to obtain corresponding encoding information; The neural network performs decoding processing according to the encoded information to obtain corresponding decoded information; The difference between the decoded information and the target centerline point sequence is calculated until the difference reaches a target value, completing the training of the neural network and obtaining the centerline extraction model.

8. The method according to claim 6 or 7, wherein before using the contour point sequence as input of a neural network and the target centerline point sequence as output of the neural network and training the neural network to obtain a centerline extraction model, the method further comprises: Normalization processing is performed on the contour point sequence and the target centerline point sequence.

9. A method for determining a guidance point, comprising: Get the target point of interest; Determining a navigation track point based on the target point of interest; The navigation track points are filtered based on the center line obtained by the method according to any one of claims 1 to 5 to obtain guide points.

10. A device for extracting the center line of a polygon, wherein: The device is applied in the field of map navigation, including: A first acquisition module is configured to acquire a polygonal contour point sequence; wherein the polygonal contour point sequence is used to describe the shape of a river or a lake; The centerline extraction module is configured to input the contour point sequence into a centerline extraction model obtained based on neural network training, the centerline extraction model outputs the centerline point sequence of the polygon, and obtains the centerline of the polygon based on the centerline point sequence.

11. The apparatus according to claim 10, further comprising: a first processing module configured to perform normalization processing on the contour point sequence before the centerline extraction module inputs the contour point sequence into the centerline extraction model; The second processing module is configured to perform inverse normalization processing on the centerline point sequence before obtaining the centerline of the polygon based on the centerline point sequence.

12. The apparatus according to claim 11, wherein the first processing module comprises: a first determining unit configured to determine a maximum value and a minimum value among coordinates of a plurality of contour points in the contour point sequence; The first calculation unit is configured to perform normalization calculation on each of the contour point coordinates based on the maximum value and the minimum value of the contour point coordinates, and limit each of the contour point coordinates to a preset value range.

13. The apparatus according to claim 11, wherein the second processing module comprises: a second determining unit configured to determine a maximum value and a minimum value among coordinates of a plurality of contour points in the contour point sequence; The second calculation unit is configured to perform an inverse normalization calculation on each center line point coordinate in the center line point sequence based on the maximum value and the minimum value of the contour point coordinates, and restore each center line point coordinate.

14. A model training device, wherein: The device is applied in the field of map navigation, including: A second acquisition module is configured to acquire a polygonal contour point sequence and a target centerline point sequence; wherein the polygonal contour point sequence is used to describe the shape of a river or a lake; The model training module is configured to use the contour point sequence as the input of the neural network and the target centerline point sequence as the output of the neural network, and train the neural network to obtain a centerline extraction model; the centerline extraction model is used to extract the centerline point sequence of the polygon.

15. The device according to claim 14, wherein The model training module includes: an input unit configured to input the contour point sequence into the neural network; An encoding unit configured to perform encoding processing on the contour point sequence by the neural network to obtain corresponding encoding information; A decoding unit configured to perform decoding processing on the neural network according to the encoded information to obtain corresponding decoded information; The calculation unit is configured to calculate the difference between the decoded information and the target centerline point sequence until the difference reaches a target value, thereby completing the training of the neural network and obtaining the centerline extraction model.

16. The apparatus according to claim 14 or 15, further comprising: The third processing module is configured as the model training module to take the contour point sequence as the input of the neural network and the target centerline point sequence as the output of the neural network, and to normalize the contour point sequence and the target centerline point sequence before training the neural network to obtain the centerline extraction model.

17. A guidance point determination device, comprising: a third acquisition module, configured to acquire a target point of interest; a determination module, configured to determine a navigation track point based on the target point of interest; The filtering module is configured to filter the navigation trajectory points based on the center line obtained by the method according to any one of claims 1 to 5 to obtain a guide point.

18. An electronic device comprising: at least one processor; as well as a memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the method according to any one of claims 1 to 9.

19. A non-transitory computer-readable storage medium storing computer instructions, wherein: The computer instructions are used to cause the computer to execute the method according to any one of claims 1 to 9.

20. A computer program product comprising a computer program, which, when executed by a processor, implements the method according to any one of claims 1 to 9.

Images