Auxiliary operation method, device and electronic device for electronic map drawing

Abstract

The present disclosure provides an auxiliary operation method, device and electronic device for electronic map drawing, which relates to the field of big data processing, and particularly to the field of electronic maps. The specific implementation solution is as follows: collect a plurality of sample data, where the sample data includes map elements and drawing areas; construct training data according to the sample data, where the training data includes map elements and feature vectors of sample images determined according to the sample data; group the plurality of training data according to the map elements to obtain a plurality of training data groups, obtain at least one training model corresponding to each training data group, and determine a drawing example corresponding to each training model; obtain a to-be-operated task, determine a training model matching the to-be-operated map elements of the to-be-operated task, and output the drawing example corresponding to the matching training model to assist in executing the to-be-operated task. The auxiliary operation method for electronic map drawing in the present disclosure can improve the accuracy of electronic map base map production and shorten the cycle of electronic map iteration.

Description

Technical Field

[0001] The present disclosure relates to the field of big data technology, and in particular, to a method, an apparatus, an electronic device, and a computer storage medium for electronic map drawing in the field of electronic maps. Background Art

[0002] In today's digital information age, through information technology means, electronic maps have gradually become the most used and most frequently used tools by people. Therefore, the quality of map data is particularly important, and improving mapping efficiency and reducing mapping costs have also become one of the most important goals. Summary of the Invention

[0003] The present disclosure provides an auxiliary operation method, an apparatus, an electronic device, and a computer storage medium for electronic map drawing.

[0004] According to one aspect of the present disclosure, there is provided an auxiliary operation method for electronic map drawing, including:

[0005] Collecting a plurality of sample data, where the sample data includes map elements and drawing areas;

[0006] Constructing corresponding training data according to the sample data, where the training data includes the map elements and feature vectors of the sample images determined according to the sample data;

[0007] Grouping a plurality of training data according to the map elements to obtain a plurality of training data groups, obtaining at least one training model corresponding to each training data group, and determining a drawing example corresponding to each training model;

[0008] Obtaining a to-be-operated task, determining a training model matching the to-be-operated map elements of the to-be-operated task, and outputting the drawing example corresponding to the matching training model to assist in executing the to-be-operated task.

[0009] According to another aspect of the present disclosure, there is provided an auxiliary operation apparatus for electronic map drawing, including:

[0010] A collection module, configured to collect a plurality of sample data, where the sample data includes map elements and drawing areas;

[0011] A construction module, configured to construct corresponding training data according to the sample data, where the training data includes the map elements and feature vectors of the sample images determined according to the sample data;

[0012] A training module, configured to group a plurality of training data according to the map elements to obtain a plurality of training data groups, obtain at least one training model corresponding to each training data group, and determine a drawing example corresponding to each training model;

[0013] An auxiliary module for obtaining a to-be-operated task, determining a training model matching a to-be-operated map element of the to-be-operated task, and outputting a drawing example corresponding to the matching training model to assist in executing the to-be-operated task.

[0014] According to another aspect of the present disclosure, there is provided an electronic device, including:

[0015] At least one processor; and

[0016] A memory communicatively connected to the at least one processor; wherein,

[0017] The memory stores instructions executable 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 execute the auxiliary operation method for electronic map drawing.

[0018] According to another aspect of the present disclosure, there is provided a non-transitory computer-readable storage medium storing computer instructions, wherein the computer instructions are used to cause the computer to execute the auxiliary operation method for electronic map drawing.

[0019] According to another aspect of the present disclosure, there is provided a computer program product including a computer program, and the computer program implements the auxiliary operation method for electronic map drawing when executed by a processor.

[0020] The auxiliary operation method for electronic map drawing according to the present disclosure can improve the accuracy of electronic map base map production, thereby shortening the cycle of electronic map iteration.

[0021] It should be understood that the content described in this part is not intended to identify the key or important features of the embodiments of the present disclosure, nor is it used to limit the scope of the present disclosure. Other features of the present disclosure will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] The drawings are used to better understand the solution and do not constitute a limitation to the present disclosure. Among them:

[0023] Figure 1 is a schematic diagram of an auxiliary operation method for electronic map drawing according to an embodiment of the present disclosure;

[0024] Figure 2 is a schematic diagram of a relationship tree between map elements and a training model according to an embodiment of the present disclosure;

[0025] Figure 3 is a schematic diagram of an auxiliary operation device for electronic map drawing according to an embodiment of the present disclosure;

[0026] Figure 4It is a block diagram of an electronic device for implementing the auxiliary operation method of electronic map drawing according to an embodiment of the present disclosure. Detailed implementation manners

[0027] The following describes exemplary embodiments of the present disclosure with reference to the accompanying drawings. Various details of the embodiments of the present disclosure are included to facilitate understanding, and they should be considered merely exemplary. Therefore, those of ordinary skill in the art should recognize that various changes and modifications can 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.

[0028] In the current process of making electronic maps, generally, several operation links such as operators' operations, quality inspections, and acceptance checks will be experienced. In this entire operation chain, the operation link is crucial, and the operation quality at this link determines the efficiency of subsequent operation links to a certain extent. If the quality inspection result of the operation link does not meet the requirements, then "rework" or "repair" operations will be carried out. Due to changes in operators, the addition of new employees, or the change of old employees' positions, and the unfamiliarity with the new work scenario, it often leads to a high number of operation errors. As a result, multiple quality inspection operations and multiple acceptance operations are required, which invisibly increases the labor input cost, prolongs the update cycle of map data iteration, and reduces the data update efficiency.

[0029] The current process of making electronic maps generally includes:

[0030] An operator manually draws on the base map according to different operation task information;

[0031] After the operator finishes drawing, a quality inspector then checks the quality according to the operation result of the operator; if the passing rate of the inspection does not meet the set value, it is sent back to the operator for re-operation, which is also called "rework" or "repair" in production. Here, rework generally refers to re-doing all operations, and repair generally only adjusts for errors. Therefore, rework brings greater input. If the passing rate of the inspection meets the set value, it is then continued to be handed over to the acceptance personnel for acceptance of the operation results.

[0032] If the acceptance result of the acceptance personnel does not meet the set value, it will continue to be returned to the operator for rework or repair operation.

[0033] The above-mentioned operation links are all completed manually. Due to the differences among operators (such as experience, level, etc.), some operators make more mistakes during operation and will be repeatedly sent back for rework, that is, they experience multiple rounds of the "operation-quality inspection-rework" link, resulting in a slow map iteration cycle and low efficiency.

[0034] To solve the above problems, an embodiment of the present disclosure provides an auxiliary operation method for electronic map drawing, such asFigure 1 As shown in Figure 1 , the method includes:

[0035] Step S101: Collect multiple pieces of sample data, where the sample data includes map elements and drawing areas.

[0036] In one example, when the map needs to be updated, a job task is generated according to the map update content. The job task includes a job area and a job object. The job area refers to where to perform the operation on the map, and the job area corresponds to the drawing area in the sample data. The drawing area can be represented by coordinate data, such as longitude and latitude. The operator can determine the job area of this job task based on the longitude and latitude.

[0037] The job object indicates what operation to perform in the job area, and the job object corresponds to the map element in the sample data. The map element is the main body that makes up the map content. According to different properties, the classification of map elements is also different. For example, when classifying map elements according to natural elements, the map elements include oceans, lakes, mountains, basins, etc.; when classifying according to traffic elements, the map elements include speed limit signs, road guiding signs, parking signs, road marking lines (such as crosswalks, turning marking lines, etc.), monitoring signs, etc. The present invention does not make specific restrictions on the classification of map elements.

[0038] After receiving the job task, the operator performs the operation according to the job task. After the operation is completed, the quality inspection personnel conduct quality inspection. The sample data is the job that has passed the quality inspection by the quality inspection personnel, that is, the data corresponding to the task where the operator has performed the operation correctly. For example, a job task is: The operator needs to draw a 30 km / h speed limit sign in the area of 25° east longitude and 20° north latitude. After the operator finishes drawing in the corresponding drawing area, the quality inspection personnel then conduct quality inspection based on the operation result of the operator. After being inspected by the quality inspection personnel, this job task is performed correctly, and the quality inspection personnel mark this job task as sample data. Therefore, a piece of sample data corresponding to this job task includes the following information: drawing area (25° east longitude, 20° north latitude), and the map element is a 30 km / h speed limit sign.

[0039] In one example, a piece of sample data can include two or more map elements. These two map elements can be of the same type or different types of map elements. For example, a piece of sample data is to draw two speed limit signs in a certain drawing area, or to draw a speed limit sign and a road guiding sign in a certain drawing area.

[0040] In one example, when an operator finishes drawing the base map of the map, a quality inspector will conduct quality inspection on the accuracy of the base map. The quality inspector checks whether the drawing area and map elements of the base map are accurate according to the content of the operation task. The quality inspector marks the base map that passes the quality inspection, that is, the correct sample data of the operation is obtained. Multiple base maps that are correct for the operation are marked according to the operation records, so as to obtain multiple sample data.

[0041] Step S102, establish corresponding training data according to the sample data.

[0042] In one example, establishing training data according to the sample data includes:

[0043] Intercept the base map according to the drawing area to obtain a sample image;

[0044] Extract features from one or more dimensions of the sample image to obtain a feature vector. The training data includes the feature vector and the map elements in this piece of sample data.

[0045] By intercepting the electronic map base map and extracting features of the sample image, the feature vector in the drawing area can be obtained.

[0046] In one example, when intercepting the base map according to the drawing area of each piece of sample data to obtain a sample image, the base map can be intercepted in a way of limiting the width and height. For example, taking the center of the drawing area corresponding to the sample data as the center point, limiting the width to 100 pixel points (pix) and the height to 100 pix as well, then the intercepted window size is 100×100 (pix), and the size of the obtained sample image is 100×100 (pix). The intercepted sizes of multiple pieces of sample data are the same. It is also possible to intercept the base map in a way of not limiting the width and height. Then, for the case where the sizes of the intercepted sample images are different, image compression technology can be used to make the sizes of all sample images the same. In this example, the intercepted window size can also be adjusted according to the drawing area and map elements of the sample data. For example, when the map element is a long and narrow road center line, the width can be limited to 200 pix and the height to 100 pix.

[0047] In one example, extracting features from one or more dimensions of the sample image to obtain a feature vector includes:

[0048] Quantize and encode the intercepted sample image according to the values of one or more dimensions of pixel points;

[0049] Then decode the quantized and encoded sample image to obtain the feature vector of the sample image.

[0050] Generally, there is a certain correlation between the pixels of an image, and there is redundant image information. By quantifying, encoding, and decoding the sample image, the redundant information can be removed to ensure the quality of the sample image.

[0051] Feature extraction is performed on one or more dimensions of the sample image, such as the gray value dimension, RGB value dimension, brightness dimension, etc. of the sample image. The present disclosure does not make specific limitations on this. In one example, taking the feature vector as a one-dimensional vector of gray values, the gray values of the sample image are extracted. The gray value of each pixel of the sample image is a value of the feature vector. For example, if the sample image is 100×100 (pix), a feature vector composed of 10,000 gray values can be obtained.

[0052] For each piece of correct sample data of a job, a corresponding training data can be obtained, and the training data includes a feature vector and a map element.

[0053] Step S103: Group the multiple pieces of training data according to the map elements to obtain multiple training data groups, obtain at least one training model corresponding to each training data group, and determine the drawing example corresponding to each training model.

[0054] Since each piece of training data contains a map element and a feature vector, the training data can be grouped according to the map elements contained in the training data. The training data with the same map elements are grouped into one group, and multiple training data groups can be obtained. In one example, a training data group can be marked, and the mark can adopt the identifier of the map element. Since the map elements contained in the training data may be one or multiple, the mark corresponding to a training data group may be the identifier of one map element or a combination of multiple map element identifiers, that is, a training data group corresponds to one map element or a map element combination.

[0055] Obtaining at least one training model corresponding to each training data group includes:

[0056] Input all the training data of a training data group into a neural network to obtain a set of weight values of an initial model, and obtain the first training model according to the set of weight values;

[0057] According to the feature vector and the first training model of each piece of training data in the training data group, calculate the eigenvalue of each piece of training data;

[0058] Divide the training data with the same eigenvalue into a training data subgroup to obtain multiple training data subgroups corresponding to the training data group;

[0059] For each subgroup of training data, all its training data is input into the neural network to obtain a set of weight values of the initial model. According to this set of weights, a second training model corresponding to each subgroup of training data is obtained;

[0060] All the second training models are used as the training models of the training data group.

[0061] By performing grouped training on the training data, training models corresponding to different training groups are obtained, improving the accuracy of model training.

[0062] In an example, assume there are 10,000 pieces of training data. Among them, the 10,000 pieces of training data are divided into four training data groups, namely training data group a, training data group b, training data group c + d, and training data group a + d. All the training data containing map element a is classified into training data group a, all the training data containing map element b is classified into training data group b, all the training data containing map elements c and d is classified into training data group c + d, and all the training data containing map elements a and d is classified into training data group a + d.

[0063] Each training data group is respectively input into the neural network, and through machine learning, at least one training model corresponding to each training group is obtained.

[0064] In an example, each piece of training data contains a feature vector composed of n gray values. The number of units in the input layer of the neural network is n + 1. In addition to the n gray values, there is also 1 offset, and the offset is a constant 1. The initial model is in polynomial form:

[0065] That is

[0066] h w (x) = W T X

[0067] where W T = [w0, w1,..., w n T ; T represents the transpose of the vector;

[0068] x0 is the offset, x0 = 1; x1,..., x n represent the n gray values that make up the feature vector in a piece of training data;

[0069] w0, w1,..., w n represent the weights corresponding to x0, x1,..., x n respectively.

[0070] ​For example, a training data set has 10,000 pieces of training data. Input these 10,000 pieces of training data into a neural network, and use the backpropagation algorithm to calculate a set of values of (w0, w1,..., w n ). Substitute the values of w0, w1,..., w n into the initial model to obtain the first training model. Input the grayscale values x0, x1,..., x n of the feature vectors of each piece of training data in the training data set into this first training model, and then the corresponding eigenvalue h w (x) of each piece of training data can be calculated.

[0071] Suppose training data 1: h w (x) = 0.995; training data 2: h w (x) = 0.994;

[0072] Training data 3: h w (x) = 0.997; training data 4: h w (x) = 0.996.

[0073] And so on, calculate the values of h w (x) for 10,000 pieces of training data. Take the training data with the same h w (x) numerical results in a training data set as a training data subgroup. Suppose after calculating 10,000 pieces of training data, 10 values of h w (x) are obtained, then this training data set is divided into 10 training data subgroups.

[0074] For each training data subgroup, input all the training data in it into the neural network, and use the backpropagation algorithm to calculate a set of (w0, w1,..., w n ). Input this set of (w0, w1,..., w n ) into the initial model to obtain the second training model corresponding to this training data subgroup. In this way, finally, 10 training models (second training models) of this training data set can be obtained. After all the training data is input into the neural network, through machine learning, finally, a relationship tree of map elements and training models as shown in Figure 2 can be obtained.

[0075] After obtaining at least one training model for each training data set, determine the drawing example for each training model. That is, select the drawing example from the sample images of all the training data in the training data subgroup corresponding to the second training model, and this drawing example is the drawing example corresponding to this second training model.

[0076] Step S104: Obtain the task to be operated, determine the training model that matches the map element to be operated in the said task to be operated, and output the drawing example corresponding to the matched training model to assist in executing the task to be operated.

[0077] In one example, the task to be operated includes the area to be drawn and the map element to be operated. Only based on the area to be drawn and the map element to be operated can an accurate match be made with the relationship tree of map elements and training models. For example, if the task to be operated is to draw a speed limit sign in the drawing area at 30° east longitude and 30° north latitude, then the area to be drawn is 30° east longitude and 30° north latitude, and the map element to be operated is the speed limit sign.

[0078] In one example, matching the training model according to the said task to be operated to obtain the drawing example that matches the task to be operated includes:

[0079] Intercept the base map according to the area to be drawn to obtain the image to be operated and acquire the feature vector of the image to be operated;

[0080] Obtain multiple training models that match the map element to be operated according to the map element to be operated in the task to be operated;

[0081] Select the training model that meets the confidence interval requirements from the multiple training models according to the feature vector of the image to be operated, and output the drawing example corresponding to this training model.

[0082] When there is a new task to be operated, through the above steps, the drawing example for the operator to refer to can be output, assisting the operator in the operation and improving the operation accuracy.

[0083] In one example, with the center of the area to be drawn as the center point, the base map is intercepted in a way of defining the width and height to obtain the image to be drawn. The size of the interception window is the same as that of the interception window of the sample image corresponding to this map element to be operated, or through image compression technology, the intercepted image to be operated has the same size as the sample image. The intercepted image to be operated is quantized and encoded according to the numerical values of one dimension or multiple dimensions of pixel points; then the quantized and encoded image to be drawn is decoded to obtain the feature vector of this image to be operated. The dimension of the feature vector of the image to be operated is the same as that of the feature vector of the sample image. If the feature vector of the sample image is a one-dimensional vector of gray values, then the feature vector of the image to be operated is also a one-dimensional vector of gray values.

[0084] In one example, input the map element to be operated in the task to be operated and traverse Figure 2The relationship tree between the map elements shown and the training models determines at least one corresponding training model according to the map elements to be processed. For example, if the map element to be processed is a speed limit sign, the training models corresponding to the speed limit sign group are the training models matched with this task to be processed; if the map element to be processed is a speed limit sign + road line, the training models corresponding to the speed limit sign + road line group are the training models matched with this task to be processed.

[0085] Then input the feature vectors of the images to be processed intercepted according to the tasks to be processed, and input the feature vectors into the multiple matching training models respectively to obtain the h w (x) values corresponding to each training model, and determine multiple h w (x) values that meet the confidence interval. w Take the h w (x) value corresponding to the training model as the training model matched with the task to be processed, and output the drawing example corresponding to this training model for the operator to refer to, so as to achieve the purpose of assisting in map drawing. For example, there are 10 training models corresponding to the map element of speed limit sign. Input the gray values x0, x1,..., x n of the feature vectors of the images to be processed into 10 training models, and 10 h w (x) values can be calculated. Assume that the 10 h w (x) values are 0.990, 0.991, 0.992, 0.993, 0.994, 0.995, 0.996, 0.997, 0.998, 0.999 respectively, and the set confidence interval is [0.994, 0.996]. Then the drawing examples of the three training models corresponding to 0.994, 0.995, and 0.996 can be output. When the operator is processing the task to be processed, the above three drawing examples can be referred to.

[0086] Adopting the auxiliary operation method for electronic map drawing in the present disclosure can improve the accuracy of electronic map base map production, thereby shortening the iteration cycle of electronic maps.

[0087] In one example, the present disclosure provides an auxiliary operation device, as Figure 3 shown. This device includes:

[0088] An acquisition module 201 for acquiring multiple pieces of sample data, where the sample data includes map elements and drawing areas;

[0089] A construction module 202 for constructing corresponding training data according to the sample data, where the training data includes the map elements and the feature vectors of the sample images determined according to the sample data;

[0090] The training module 203 is configured to group multiple pieces of training data according to the map elements, obtain multiple groups of training data, acquire at least one training model corresponding to each group of training data, and determine a drawing example corresponding to each training model.

[0091] The auxiliary module 204 is configured to obtain a to-be-operated task, determine a training model that matches the to-be-operated map elements of the to-be-operated task, and output the drawing example corresponding to the matching training model to assist in executing the to-be-operated task.

[0092] The auxiliary operation device for electronic map drawing of the present disclosure can improve the accuracy of making the base map of the electronic map, thereby shortening the iteration cycle of the electronic map.

[0093] In one example, the construction module 202 is specifically configured to:

[0094] Intercept a sample image from the base map of the electronic map according to the drawing area;

[0095] Extract features of a set dimension of the sample image to obtain a feature vector.

[0096] In one example, the training module 203 is specifically configured to:

[0097] Divide the training data with the same map elements into one group of training data; one map element or combination of map elements corresponds to each group of training data.

[0098] In one example, the training module 203 is further specifically configured to:

[0099] Input all the training data of the training data group into a neural network to obtain a set of weight values of an initial model, and obtain a first training model according to this set of weight values;

[0100] Calculate the eigenvalue of each piece of training data according to the feature vector of each piece of training data in the training data group and the first training model;

[0101] Divide the training data with the same eigenvalue into one subgroup of training data to obtain multiple subgroups of training data corresponding to the training data group;

[0102] For each subgroup of training data, input all its training data into a neural network to obtain a set of weight values of an initial model, and obtain a second training model corresponding to each subgroup of training data according to this set of weights;

[0103] Use all the second training models as the training models of the training data group.

[0104] In one example, the to-be-operated task includes a to-be-drawn area and the to-be-operated map elements.

[0105] In one example, the auxiliary module 204 is specifically configured to:

[0106] Intercept a to-be-operated image from the base map of the electronic map according to the to-be-drawn area, and obtain a feature vector of the to-be-operated image;

[0107] Determine at least one second training model that matches the to-be-operated map element;

[0108] Select at least one second training model that meets the confidence interval requirement from all the matched second training models according to the feature vector of the to-be-operated image, and output a drawing sample corresponding to the selected second training model.

[0109] According to an embodiment of the present disclosure, the present disclosure further provides an electronic device, a readable storage medium, and a computer program product.

[0110] In one example, the present disclosure provides a non-transitory computer-readable storage medium storing computer instructions, wherein the computer instructions are used to cause the computer to execute the auxiliary operation method for electronic map drawing.

[0111] In one example, the present disclosure provides a computer program product including a computer program, and the computer program implements the auxiliary operation method for electronic map drawing when executed by a processor

[0112] In one example, the present disclosure provides an electronic device, including:

[0113] At least one processor; and

[0114] A memory communicatively connected to the at least one processor; wherein,

[0115] The memory stores instructions executable 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 execute the auxiliary operation method for electronic map drawing.

[0116] Figure 4 FIG. shows a schematic block diagram of an example electronic device 300 that can be used to implement embodiments of the present disclosure. The electronic device is intended to represent various forms of digital computers, such as, a laptop computer, a desktop computer, a workbench, a personal digital assistant, a server, a blade server, a mainframe computer, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as, a personal digital processor, a cellular phone, a smart phone, a wearable device, 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.

[0117] As Figure 4 shown, device 300 includes a computing unit 301 that can perform various appropriate actions and processes according to a computer program stored in a read-only memory (ROM) 302 or a computer program loaded from a storage unit 308 into a random access memory (RAM) 303. In the RAM 303, various programs and data required for the operation of device 300 can also be stored. The computing unit 301, the ROM 302, and the RAM 303 are connected to each other via a bus 304. An input / output (I / O) interface 305 is also connected to the bus 304.

[0118] Multiple components in device 300 are connected to the I / O interface 305, including: an input unit 306, such as a keyboard, a mouse, etc.; an output unit 307, such as various types of displays, speakers, etc.; a storage unit 308, such as a magnetic disk, an optical disc, etc.; and a communication unit 309, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 309 allows device 300 to exchange information / data with other devices via a computer network such as the Internet and / or various telecommunication networks.

[0119] The computing unit 301 can be various general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of the computing unit 301 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 running machine learning model algorithms, a digital signal processor (DSP), and any appropriate processor, controller, microcontroller, etc. The computing unit 301 executes the various methods and processes described above, such as an auxiliary operation method for electronic map drawing. For example, in some embodiments, the auxiliary operation method for electronic map drawing can be implemented as a computer software program that is tangibly contained in a machine-readable medium, such as the storage unit 308. In some embodiments, part or all of the computer program can be loaded and / or installed onto device 300 via the ROM 302 and / or the communication unit 309. When the computer program is loaded into the RAM 303 and executed by the computing unit 301, one or more steps of the auxiliary operation method for electronic map drawing described above can be executed. Alternatively, in other embodiments, the computing unit 301 can be configured to execute the auxiliary operation method for electronic map drawing in any other appropriate manner (e.g., by means of firmware).

[0120] The various embodiments of the systems and techniques described above in this specification can be implemented in digital electronic circuitry, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chip (SOCs), complex 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 interpretable on a programmable system including at least one programmable processor, which may be a special-purpose or general-purpose programmable processor that receives data and instructions from, and transmits data and instructions to, a storage system, at least one input device, and at least one output device.

[0121] The program code for implementing the methods 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, special purpose computer, or other programmable data processing apparatus, such that the program codes, when executed by the processor or controller, cause the functions / operations specified in the flowchart and / or block diagram to be implemented. The program code can be executed entirely on the machine, partly on the machine, as a stand-alone software package partly on the machine and partly on a remote machine, or entirely on the remote machine or server.

[0122] In the context of the present disclosure, a machine-readable medium may be a tangible medium that can contain or store a program for use by or in connection with an instruction execution system, apparatus, or device. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of a machine-readable storage medium would include an electrical connection based on one or more wires, a portable computer diskette, 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 disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0123] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device for displaying information to the user (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor); and a keyboard and a pointing device (e.g., a mouse or a trackball) through which the user can provide input to the computer. Other kinds 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).

[0124] The systems and techniques described herein can be implemented in a computing system including backend components (e.g., as a data server), or a computing system including middleware components (e.g., an application server), or a computing system including frontend components (e.g., a user computer having a graphical user interface or a web browser through which the user can interact with an implementation of the systems and techniques described herein), or a computing system including any combination of such backend components, middleware components, or frontend components. The components of the system can be interconnected by digital data communication in any form or medium (e.g., a communication network). Examples of communication networks include: local area network (LAN), wide area network (WAN), and the Internet.

[0125] A computer system can include a client and a server. The client and the server are generally remote from each other and typically interact through a communication network. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship with each other. The server can be a cloud server, a server of a distributed system, or a server incorporating a blockchain.

[0126] It should be understood that various forms of the processes shown above can be used, with steps reordered, added, or deleted. For example, the steps recited in this disclosure can be executed 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, and no limitations are imposed herein.

[0127] The above specific embodiments do not constitute a limitation on the protection scope of this disclosure. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to 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 protection scope of this disclosure.

Claims

1. An auxiliary operation method for electronic map drawing, comprising: Collecting multiple pieces of sample data, where the sample data includes map elements and a drawing area; Constructing corresponding training data according to the sample data, where the training data includes the map elements and feature vectors of the sample images determined according to the sample data; Grouping multiple pieces of training data according to the map elements to obtain multiple training data groups, obtaining at least one training model corresponding to each training data group, and determining a drawing example corresponding to each training model; Obtaining a to-be-operated task, determining a training model that matches the to-be-drawn map elements of the to-be-operated task, and outputting the drawing example corresponding to the matched training model to assist in executing the to-be-operated task; Wherein, constructing the corresponding training data according to the sample data includes: intercepting a sample image from the base map of the electronic map according to the drawing area; extracting features from a set dimension of the sample image to obtain feature vectors.

2. The method according to claim 1, wherein, The extracting features from a set dimension of the sample image to obtain feature vectors includes: Performing quantization encoding on the values of the set dimension of each pixel point in the sample image; Decoding the quantized-encoded sample image to obtain the feature vector of the sample image.

3. The method according to claim 1, wherein, The grouping multiple pieces of training data according to the map elements includes: Dividing the training data with the same map elements into one training data group; each training data group corresponds to one map element or a combination of map elements.

4. The method according to claim 1, wherein Obtaining at least one training model corresponding to each training data group includes: Inputting all the training data of the training data group into a neural network to obtain a set of weight values of an initial model, and obtaining a first training model according to the set of weight values; Calculating the feature values of each piece of training data according to the feature vectors of each piece of training data in the training data group and the first training model; Dividing the training data with the same feature values into one training data subgroup to obtain multiple training data subgroups corresponding to the training data group; For each training data subgroup, inputting all its training data into a neural network to obtain a set of weight values of an initial model, and obtaining a second training model corresponding to each training data subgroup according to the set of weights; Taking all the second training models as the training models of the training data group.

5. The method according to claim 4, wherein, Determining a drawing example corresponding to each training model includes: Selecting a drawing example from the sample images of all the training data in the training data subgroup corresponding to the second training model.

6. The method according to claim 5, wherein The to-be-operated task includes a to-be-drawn area and the to-be-drawn map elements.

7. The method according to claim 6, wherein Determining a training model that matches the to-be-drawn map elements of the to-be-operated task and outputting the drawing example corresponding to the matched training model includes: Intercepting a to-be-operated image from the base map of the electronic map according to the to-be-drawn area, and obtaining the feature vector of the to-be-operated image; Determining at least one second training model that matches the to-be-drawn map elements; Selecting at least one second training model that meets the confidence interval requirement from all the matched second training models according to the feature vector of the to-be-operated image, and outputting the drawing example corresponding to the selected second training model.

8. An auxiliary operation device for electronic map drawing, comprising: A collection module for collecting multiple pieces of sample data, where the sample data includes map elements and drawing areas; A construction module for constructing corresponding training data according to the sample data, where the training data includes the map elements and feature vectors of the sample images determined according to the sample data; A training module for grouping multiple pieces of training data according to the map elements to obtain multiple training data groups, obtaining at least one training model corresponding to each training data group, and determining a drawing example corresponding to each training model; An auxiliary module for obtaining a to-be-operated task, determining a training model matching the to-be-operated map elements of the to-be-operated task, and outputting the drawing example corresponding to the matching training model to assist in executing the to-be-operated task; Wherein, the construction module is specifically used for: intercepting a sample image from the base map of the electronic map according to the drawing area; performing feature extraction on the set dimension of the sample image to obtain a feature vector.

9. The apparatus according to claim 8, wherein, The training module is specifically used for: Dividing the training data with the same map elements into one training data group; one map element or combination of map elements corresponding to each training data group.

10. The apparatus according to claim 9, wherein, The training module is specifically further used for: Inputting all the training data of the training data group into a neural network to obtain a set of weight values of an initial model, and obtaining a first training model according to this set of weight values; Calculating the feature values of each piece of training data according to the feature vectors of each piece of training data in the training data group and the first training model; Dividing the training data with the same feature values into one training data subgroup to obtain multiple training data subgroups corresponding to the training data group; For each training data subgroup, inputting all its training data into a neural network to obtain a set of weight values of an initial model, and obtaining a second training model corresponding to each training data subgroup according to this set of weights; Taking all the second training models as the training models of the training data group.

11. The device according to claim 8, wherein, The to-be-operated task includes a to-be-drawn area and the to-be-operated map elements.

12. The apparatus according to claim 11, wherein, The auxiliary module is specifically used for: Intercepting a to-be-operated image from the base map of the electronic map according to the to-be-drawn area, and obtaining the feature vector of the to-be-operated image; Determining at least one second training model matching the to-be-operated map elements; Selecting at least one second training model that meets the confidence interval requirement from all the matching second training models according to the feature vector of the to-be-operated image, and outputting the drawing example corresponding to the selected second training model.

13. An electronic device, 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 the method according to any one of claims 1-7.

14. 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 1-7.

15. A computer program product comprising a computer program which, when executed by a processor, implements the method according to any one of claims 1 - 7.

Images