Truck parking area identification method and device, equipment and storage medium

Abstract

The invention discloses a truck parking area identification method and device, equipment and a storage medium. The method comprises the steps of performing spatial indexing based on truck parking data in a preset time period to obtain an H3 grid corresponding to a truck parking point; performing statistical analysis based on the H3 grid identifier to obtain a high-heat H3 grid; based on the grid center coordinate corresponding to the high-heat H3 grid, capturing a remote sensing image map of a parking area and a remote sensing image map of the periphery of the parking area; and inputting the remote sensing image into a pre-trained truck parking area identification model, and outputting an identification result of the truck parking area. The invention relates to the field of intelligent traffic freight big data, identifies the position of a public parking area of a truck based on vehicle driving track point data and a remote sensing image map, can accurately and quickly identify the parking area of the truck based on truck track data, and is used for improving the efficiency of marking a fence of the parking area of the truck by marking personnel.

Description

Technical Field

[0001] The present application relates to the technical field of traffic big data, and in particular to a method, device, equipment and storage medium for identifying a truck parking area. Background Art

[0002] At present, it is a common method to annotate fences by purchasing third-party parking area POI (point of interest) data. However, this method has some limitations. First, third-party POI data includes non-truck parking areas, which leads to a waste of annotation personnel costs because these irrelevant data need to be processed. Secondly, third-party POI data cannot fully cover all truck parking areas, which reduces the effectiveness of the data, making it possible for fence annotations based on these data to not accurately reflect the actual truck parking needs and distribution. These problems point out the shortcomings of existing technologies in truck parking area identification, and also highlight the need to develop more efficient and accurate truck parking area identification technologies. Summary of the invention

[0003] The embodiments of the present application provide a method, device, equipment and storage medium for identifying a truck parking area, so as to at least solve the technical problem in the related art that it is difficult to accurately and comprehensively identify a truck parking area.

[0004] According to one aspect of an embodiment of the present application, a method for identifying a truck parking area is provided, comprising:

[0005] Spatial indexing is performed based on the truck stop data within a preset time period to obtain the H3 grid corresponding to the truck stop point;

[0006] Based on the H3 grid identification, statistical analysis was performed to obtain the high-heat H3 grid;

[0007] Based on the grid center coordinates corresponding to the high-heat H3 grid, a remote sensing image of the docking area and its surroundings is captured;

[0008] The remote sensing image is input into a pre-trained truck parking area recognition model, and a recognition result of the truck parking area is output.

[0009] In one embodiment, before performing spatial indexing based on truck stop data within a preset time period, the method further includes:

[0010] Perform stop analysis on the truck trajectory data within a preset time period to obtain the stop start point and stop end point;

[0011] Calculate the stop duration based on the stop start point and the stop end point;

[0012] The coordinates of the stop point are obtained based on the average values ​​of the horizontal coordinates and the vertical coordinates of the stop start point coordinates and the stop end point coordinates;

[0013] The truck stop data is obtained based on the stop duration, stop point coordinates and vehicle identification.

[0014] In one embodiment, after obtaining the truck stop data, the method further includes:

[0015] Based on the matching of the stop point coordinates with the map data, obtaining the point of interest information corresponding to the stop point coordinates;

[0016] Based on the information of points of interest corresponding to the coordinates of the stop points, filtering is performed to filter out the stop points in the public parking area that are not trucks, and to obtain filtered truck stop data.

[0017] In one implementation, statistical analysis is performed based on the H3 grid identifier to obtain a high-heat H3 grid, including:

[0018] All truck stops are aggregated according to the H3 grid identification;

[0019] The stop frequency and number of parked vehicles in each grid are counted based on the H3 grid identification;

[0020] The H3 grids whose stop frequency is greater than a preset first threshold and whose number of parked vehicles is greater than a preset second threshold are screened out to obtain the high-heat H3 grid.

[0021] In one implementation, based on the grid center coordinates corresponding to the high-heat H3 grid, capturing a remote sensing image of the docking area and the surrounding area of ​​the docking area includes:

[0022] Based on the center coordinates of the grid, a remote sensing image corresponding to the docking area is captured as a center remote sensing image;

[0023] Taking the center coordinates of the grid as the center, obtain the remote sensing image pictures of the surrounding eight H3 grids as the surrounding remote sensing image map;

[0024] The central remote sensing image and the peripheral remote sensing image are spliced ​​into a nine-square remote sensing image.

[0025] In one embodiment, before inputting the remote sensing image into a pre-trained truck parking area recognition model, the method further includes:

[0026] Obtain multiple remote sensing images of the docking area and its surrounding areas, and annotate them to obtain a training set and a test set;

[0027] Use the wide residual network model to perform multiple iterations of training on the training set, and generate a model file for each iteration;

[0028] Evaluate the model files on the test set and calculate the precision and recall of each model file by comparing the model output with the actual labels;

[0029] Based on the accuracy and recall rate, the model file with the best performance is selected as the trained truck parking area recognition model.

[0030] In one embodiment, the remote sensing image is input into a pre-trained truck parking area recognition model, and the truck parking area recognition result is output, including:

[0031] The remote sensing image is input into a pre-trained truck parking area recognition model, and the recognition result of the truck parking area, the corresponding H3 grid mark and the grid center coordinates are output.

[0032] According to another aspect of the embodiment of the present application, there is also provided a device for identifying a truck parking area, comprising:

[0033] The stop analysis module is used to perform spatial indexing based on the truck stop data within a preset time period to obtain the H3 grid corresponding to the truck stop point;

[0034] The screening module is used to perform statistical analysis based on H3 grid identifiers to obtain high-heat H3 grids;

[0035] A remote sensing image acquisition module is used to capture the remote sensing image of the docking area and the surrounding area based on the grid center coordinates corresponding to the high-heat H3 grid;

[0036] The recognition module is used to input the remote sensing image into a pre-trained truck parking area recognition model and output a recognition result of the truck parking area.

[0037] According to another aspect of the embodiment of the present application, there is also provided an electronic device, including a memory and a processor, wherein the memory stores a computer program, and the processor is configured to execute the above-mentioned method for identifying a truck parking area through the computer program.

[0038] According to another aspect of the embodiment of the present application, a computer-readable storage medium is provided, in which a computer program is stored, wherein the computer program is configured to execute the above-mentioned truck parking area identification method when running.

[0039] The technical solution provided by the embodiments of the present application may have the following beneficial effects:

[0040] The present application relates to the field of intelligent transportation freight big data. Based on the driving trajectory point data of freight vehicles and remote sensing images, the remote sensing images are input into the parking area recognition model to identify the location of the truck parking area. The solution of the present application can automatically identify the truck parking area by utilizing the vehicle driving trajectory point data and remote sensing images, significantly improving the efficiency of labeling personnel in marking the fences of truck parking areas. This technology saves the time and energy of labelers by reducing the erroneous labeling of non-truck parking areas, thereby reducing costs. The method of the present application also improves the coverage and effectiveness of the data, and better serves the field of intelligent transportation freight big data. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0042] Figure 1 is a flow chart of a method for identifying a truck parking area according to an embodiment of the present application;

[0043] Figure 2 is a schematic diagram of a method for constructing a truck stop point dataset according to an embodiment of the present application;

[0044] Figure 3 is a schematic diagram of a model training method according to an embodiment of the present application;

[0045] Figure 4 is a schematic diagram of a method for identifying a truck parking area by using a model according to an embodiment of the present application;

[0046] Figure 5 is a schematic diagram of a truck parking area identification device according to an embodiment of the present application;

[0047] Figure 6 It is a schematic diagram of the structure of an optional electronic device according to an embodiment of the present application. DETAILED DESCRIPTION

[0048] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present application.

[0049] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0050] The following is combined with Figure 1-6 The method for identifying a truck parking area in an embodiment of the present application is described in detail. It should be noted that the following application scenarios are only shown to facilitate understanding of the spirit and principle of the present application, and the implementation of the present application is not limited in this regard. On the contrary, the implementation of the present application can be applied to any applicable scenario. Figure 1 As shown, the method mainly includes the following steps:

[0051] S101 performs spatial indexing based on the truck stop data within a preset time period to obtain an H3 grid corresponding to the truck stop point.

[0052] In one embodiment, before performing spatial indexing based on the truck stop data within a preset time period, the method further includes acquiring the track data of the truck, and obtaining the truck stop data based on the truck track data.

[0053] Specifically, the truck trajectory data within a preset time period is analyzed for stopovers to obtain stop start points and stop end points.

[0054] Select the historical truck trajectory data within a time range of t1 to t2. Analyze the vehicle's stop start point. In one embodiment, if the speed of two consecutive points is less than v kilometers per hour, and the duration is less than or equal to t seconds, then the first point is the stop start point. Alternatively, when the time difference between two consecutive points is greater than t seconds, and the distance between the two points is less than s meters, the speed of the first point is less than v kilometers per hour, and the first point is the stop start point. Among them, the specific values ​​of the preset thresholds v, t, and s can be set according to actual conditions, and the embodiments of the present application are not limited.

[0055] Further, the vehicle's stop end point is analyzed. In one embodiment, if there are consecutive points with a speed greater than or equal to v kilometers per hour, the time difference between the first point and the last point with a speed less than v kilometers per hour is less than t seconds, and the distance is less than s meters, the first point greater than or equal to v kilometers per hour is taken as the stop end point. Otherwise, the point before the first point with a speed greater than or equal to v kilometers per hour is taken as the stop end point. Among them, the specific values ​​of the preset thresholds v, t, and s can be set according to actual conditions, and the embodiments of the present application do not limit them.

[0056] Further, based on the stop start point and the stop end point, the stop duration is calculated. According to the time corresponding to the stop start point and the stop end point, the stop duration = stop end time - stop start time.

[0057] Further, the stop point coordinates are obtained based on the average values ​​of the abscissa and ordinate of the stop start point coordinates and the stop end point coordinates.

[0058] In one embodiment, the center coordinates are calculated based on the docking start coordinates and the docking end coordinates, and the docking point coordinates are marked. The starting coordinates are A(x,y), the ending coordinates are B(x1,y1), and the docking point coordinates are: o((x+x1) / 2, (y+y1) / 2).

[0059] Based on the stop duration, stop point coordinates and vehicle identification, the truck stop data is obtained.

[0060] In one implementation, all trajectory data from the time range t1 to t2 are used to calculate all stop points, and the coordinates, stop duration, and vehicle ID of each stop point are output.

[0061] Furthermore, based on the matching of the stop point coordinates with the map data, the point of interest information corresponding to the stop point coordinates is obtained; based on the point of interest information corresponding to the stop point coordinates, filtering is performed to filter out the stop points in non-truck public parking areas to obtain filtered truck stop data.

[0062] Specifically, detailed map data is obtained, including point of interest (POI) information for gas stations, roads, service areas, manufacturing enterprises, logistics parks, and living areas. The coordinates of truck stops are matched with map data using geographic information system (GIS) technology to determine the geographic location of each stop. Using the POI information in the map data, specific areas that match the stops, such as gas stations, service areas, etc., are identified.

[0063] Furthermore, non-target areas are filtered: based on the matching results, stops located in non-truck parking areas (such as gas stations, service areas, logistics parks, manufacturing enterprises, etc.) are filtered out. The filtered stop data is output, which represents the parking situation of trucks in public parking areas.

[0064] After obtaining the filtered truck stop data, spatial indexing is performed to obtain the H3 grid corresponding to the truck stop point.

[0065] In the implementation of the present application, the truck stop data includes latitude and longitude information. Then select H3 resolution: determine the appropriate H3 grid resolution (level). The lower the resolution, the larger the geographical area covered by each grid; the higher the resolution, the finer the grid. Use functions in the H3 library, such as ST_H3FromLatLng, to convert the latitude and longitude coordinates of each stop point into H3 code. This function requires the longitude and latitude and the target level to be specified, and the corresponding H3 code can be directly obtained.

[0066] The converted H3 code is used as a spatial index. Each code represents a unique hexagonal grid unit, so that the truck stop data is indexed into the H3 grid system.

[0067] S102 performs statistical analysis based on the H3 grid identifier to obtain a high-heat H3 grid.

[0068] In one embodiment, a high-heat H3 grid is obtained by performing statistical analysis based on the H3 grid identifier, including: aggregating all truck stops according to the H3 grid identifier; counting the stop frequency and the number of parked vehicles for each grid based on the H3 grid identifier; and screening out H3 grids whose stop frequency is greater than a preset first threshold and whose number of parked vehicles is greater than a preset second threshold to obtain a high-heat H3 grid.

[0069] Specifically, all stops are grouped and aggregated according to their H3 grid identifiers. Statistics are taken for each group of H3 grid identifiers to calculate the stop frequency of each grid (i.e., the number of all stops in the grid). At the same time, the number of parked vehicles in each grid (i.e., the number of all different vehicle IDs in the grid) is calculated.

[0070] Determine two preset thresholds: the first threshold is for the frequency of stops, and the second threshold is for the number of stopped vehicles. According to the statistical results, filter out those H3 grids whose stop frequency is greater than the first threshold and whose number of stopped vehicles is greater than the second threshold. Output the list of filtered high-heat H3 grids, which represent the hot spots for truck stops. Among them, the setting of specific thresholds is set according to actual needs, and this application does not make specific restrictions.

[0071] S103 captures the remote sensing image of the docking area and its surroundings based on the grid center coordinates corresponding to the high-heat H3 grid.

[0072] In one implementation, first, based on the center coordinates of the grid, a remote sensing image corresponding to the parking area is captured as the center remote sensing image.

[0073] Among them, the grid center coordinates can be obtained based on the stop point coordinates in the H3 grid. According to the grid center coordinates, the corresponding central remote sensing image map is captured through the remote sensing image service API.

[0074] Furthermore, taking the grid center coordinates as the center, remote sensing image pictures of eight surrounding H3 grids are obtained as the surrounding remote sensing image map.

[0075] Get the surrounding H3 grids. For each surrounding H3 grid identifier, also grab the corresponding remote sensing image through the remote sensing image service API.

[0076] The central remote sensing image and the surrounding remote sensing image are spliced ​​into a nine-square remote sensing image.

[0077] The central remote sensing image and the eight surrounding remote sensing images are spliced ​​in a nine-square grid layout. Image processing software or programming libraries can be used to achieve image splicing. The spliced ​​nine-square grid image is saved and the image is named as the central H3 grid logo. The size is 256*256. By splicing the surrounding remote sensing images, the truck parking area can be better described and the recognition accuracy can be improved.

[0078] S104 inputs the remote sensing image into a pre-trained truck parking area recognition model and outputs a recognition result of the truck parking area.

[0079] In one embodiment, before inputting the remote sensing image into the pre-trained truck parking area recognition model, the method further includes:

[0080] Multiple remote sensing images of the docking area and its surrounding areas are obtained and annotated to obtain a training set and a test set.

[0081] Obtain multiple remote sensing images of the docking area and its surroundings. The remote sensing images at this time are spliced ​​nine-square remote sensing images. Then perform manual labeling, and mark the results as "yes" or "no". The labeled data set is randomly divided into P1 training set and P2 test set with the test set accounting for 20%. The specific ratio can be adjusted according to needs.

[0082] Furthermore, the wide residual network model is used to perform multiple iterations of training on the training set, and a model file is generated for each iteration.

[0083] The images of the P1 training set are used as input and trained through the WideResNet50_2, a wide residual network model. This model uses 50 convolutional layers and a wide factor with a large number of parameters, providing stronger image feature extraction and better image classification effects. It iterates N cycles, traversing the entire data set each time, and using all samples in the training set once to generate model files for each iteration, totaling N.

[0084] Evaluate the model files on the test set and calculate the precision and recall of each model file by comparing the model output with the actual labels.

[0085] The images of the P2 test set are used as input, and N model files are loaded cyclically through the WideResNet50_2 (wide residual network) algorithm to calculate the accuracy and recall of each model file. Recognize a picture, output "yes" or "no", and compare it with the marked label. If it is consistent, record the model file true value +1, if it is consistent and the label is "yes", record the model file recall value +1, each model file accuracy = all true values ​​and / test set number, recall = recall value and / number of marked labels "yes".

[0086] Based on the accuracy and recall rate, the model file with the best performance is selected as the trained truck parking area recognition model.

[0087] In an optional implementation, an accuracy threshold Q1 and a recall threshold Q2 are set. A model file with an accuracy greater than Q1 and a recall greater than Q2 is selected as the optimal model file output. The specific threshold values ​​are not limited.

[0088] If there are multiple models that meet the threshold conditions, a weighted sum is taken based on the precision and recall rates, and the models are sorted from high to low based on the weighted sum values. The model file ranked first is taken as the optimal output.

[0089] After obtaining the trained truck parking area recognition model, the remote sensing image to be recognized is input into the pre-trained truck parking area recognition model, and the recognition result of the truck parking area is output, and the recognition result is "yes" or "no". The corresponding H3 grid identifier and grid center coordinates are output.

[0090] In order to facilitate understanding of the truck parking area identification method of the present application embodiment, the following is combined with the attached Figure 2-4 Further description.

[0091] Figure 2 A schematic diagram of a method for constructing a truck stop dataset.

[0092] The method comprises: S201: selecting a trajectory to calculate a stop point; selecting and calculating possible stop points from the vehicle's driving trajectory data.

[0093] S202: Filter out the stops in non-public parking areas; from the calculated stops, filter out those that do not belong to truck parking areas, such as logistics parks, service areas, and other areas.

[0094] S203: Generate H3 grid IDs from the stop points; convert the remaining truck parking area stop points into H3 grid IDs, where H3 is a geospatial indexing system that divides the earth's surface into hexagonal grids.

[0095] S204: Select high-heat H3 grids; from the generated H3 grid IDs, select those grids with high parking frequency or large number of vehicles, which are considered to be high-heat parking areas.

[0096] S205: Output grid ID, frequency, number of vehicles, grid center coordinates; finally, output the ID, parking frequency, number of vehicles involved, and grid center coordinates of each high-heat H3 grid. The entire flowchart describes an automated processing process from raw trajectory data to the final output of high-heat parking area information.

[0097] Figure 3 It is a schematic diagram of a model training method according to an embodiment of the present application.

[0098] It includes: S301: collecting image maps based on the stop points of the data set; using the data of the vehicle stop points, collecting corresponding remote sensing image maps or map images, and these images will be used for subsequent analysis and model training.

[0099] S302: Filter N sample sets from the data set and annotate them: Filter N samples from the collected image data and annotate them manually. These samples will be used to train and test the model.

[0100] S303: The labeled output data includes the grid ID and whether it is a parking area; the screened samples are labeled, and the output data should include the H3 grid ID of each sample and information on whether the area is a parking area.

[0101] S304: Generate training and test data sets; according to the annotation results, divide the data sets into training sets and test sets for model training and performance evaluation.

[0102] S305: training a model based on a wide residual network algorithm; using a wide residual network algorithm and a training data set, training a model to identify parking areas. A wide residual network is a deep learning model that improves the performance of the model by increasing the width of the network and is suitable for image classification tasks.

[0103] The entire flowchart describes a complete process from data collection, sample screening, data labeling, data set division to model training, aiming to automatically identify and analyze vehicle parking areas through machine learning technology.

[0104] Figure 4 The present invention is a schematic diagram of a method for identifying a truck parking area through a model according to an embodiment of the present invention, comprising: S401: inputting a remote sensing image to be identified into the model; preparing remote sensing images to be identified, and providing these images as input data to a pre-trained model.

[0105] S402: Identify parking areas based on the model: Use the trained model to analyze the input remote sensing image to identify parking areas. This step is the key to the application of the model, which will predict whether the image contains a parking area based on the learned features.

[0106] S403: Output the H3 grid ID, center coordinates and whether it is parked in the public area; after model recognition, output the H3 grid ID, grid center coordinates and an identifier corresponding to each parking area to indicate whether the area is a truck parking area.

[0107] The solution of this application can automatically identify the coordinates of truck parking areas by using vehicle driving trajectory point data and remote sensing images, significantly improving the efficiency of labeling personnel in marking truck parking area fences. This technology saves labeling personnel's time and energy by reducing incorrect labeling of non-truck parking areas, thereby reducing costs. In addition, the method of this application improves the coverage and effectiveness of data, and better serves the field of intelligent transportation freight big data.

[0108] According to another aspect of the embodiment of the present application, a truck parking area identification device for implementing the above truck parking area identification method is also provided. Figure 5 As shown, the device comprises:

[0109] The stop analysis module 501 is used to perform spatial indexing based on the truck stop data within a preset time period to obtain the H3 grid corresponding to the truck stop point;

[0110] A screening module 502 is used to perform statistical analysis based on the H3 grid identifier to obtain a high-heat H3 grid;

[0111] The remote sensing image acquisition module 503 is used to capture the remote sensing image of the docking area and the surrounding area based on the grid center coordinates corresponding to the high-heat H3 grid;

[0112] The recognition module 504 is used to input the remote sensing image into the pre-trained truck parking area recognition model and output the recognition result of the truck parking area.

[0113] It should be noted that the truck parking area identification device provided in the above embodiment only uses the division of the above functional modules as an example when executing the truck parking area identification method. In actual applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above. In addition, the truck parking area identification device provided in the above embodiment and the truck parking area identification method embodiment belong to the same concept, and the implementation process thereof is detailed in the method embodiment, which will not be repeated here.

[0114] According to another aspect of the embodiments of the present application, an electronic device corresponding to the truck parking area identification method provided in the above embodiments is also provided to execute the above truck parking area identification method.

[0115] Please refer to Figure 6 , which shows a schematic diagram of an electronic device provided by some embodiments of the present application. Figure 6 As shown, the electronic device includes: a processor 600, a memory 601, a bus 602 and a communication interface 603, and the processor 600, the communication interface 603 and the memory 601 are connected via the bus 602; the memory 601 stores a computer program that can be run on the processor 600, and when the processor 600 runs the computer program, it executes the truck parking area identification method provided in any of the aforementioned embodiments of the present application.

[0116] The memory 601 may include a high-speed random access memory (RAM), and may also include a non-volatile memory, such as at least one disk memory. The communication connection between the system network element and at least one other network element is realized through at least one communication interface 603 (which may be wired or wireless), and the Internet, wide area network, local area network, metropolitan area network, etc. may be used.

[0117] The bus 602 may be an ISA bus, a PCI bus, or an EISA bus, etc. The bus may be divided into an address bus, a data bus, a control bus, etc. The memory 601 is used to store programs, and the processor 600 executes the programs after receiving the execution instructions. The method for identifying a truck parking area disclosed in any of the embodiments of the present application may be applied to the processor 600, or implemented by the processor 600.

[0118] The processor 600 may be an integrated circuit chip with signal processing capabilities. In the implementation process, each step of the above method can be completed by the hardware integrated logic circuit or software instructions in the processor 600. The above processor 600 can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components. The methods, steps and logic block diagrams disclosed in the embodiments of the present application can be implemented or executed. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor. The steps of the method disclosed in the embodiments of the present application can be directly embodied as a hardware decoding processor to be executed, or the hardware and software modules in the decoding processor can be executed. The software module can be located in a mature storage medium in the field such as a random access memory, a flash memory, a read-only memory, a programmable read-only memory or an electrically erasable programmable memory, a register, etc. The storage medium is located in the memory 601, and the processor 600 reads the information in the memory 601 and completes the steps of the above method in combination with its hardware.

[0119] The electronic device provided in the embodiment of the present application and the method for identifying the truck parking area provided in the embodiment of the present application are based on the same inventive concept and have the same beneficial effects as the methods adopted, operated or implemented therein.

[0120] According to another aspect of the embodiments of the present application, there is also provided a computer-readable storage medium corresponding to the method for identifying a truck parking area provided in the aforementioned embodiments, on which a computer program (i.e., a program product) is stored. When the computer program is executed by a processor, it will execute the method for identifying a truck parking area provided in any of the aforementioned embodiments.

[0121] It should be noted that examples of computer-readable storage media may also include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other optical or magnetic storage media, which are not listed here one by one.

[0122] The computer-readable storage medium provided in the above-mentioned embodiment of the present application and the method for identifying a truck parking area provided in the embodiment of the present application are based on the same inventive concept and have the same beneficial effects as the method adopted, run or implemented by the application program stored therein.

[0123] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0124] The above embodiments only express several implementation methods of the present invention, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present invention. It should be pointed out that, for a person of ordinary skill in the art, several modifications and improvements can be made without departing from the concept of the present invention, and these all belong to the protection scope of the present invention. Therefore, the protection scope of the present invention patent shall be subject to the attached claims.

Claims

1. A method for identifying a truck parking area, characterized in that: include: Spatial indexing is performed based on the truck stop data within a preset time period to obtain the H3 grid corresponding to the truck stop point; Based on the H3 grid identification, statistical analysis was performed to obtain the high-heat H3 grid; Based on the grid center coordinates corresponding to the high-heat H3 grid, a remote sensing image of the docking area and its surroundings is captured; The remote sensing image is input into a pre-trained truck parking area recognition model, and a recognition result of the truck parking area is output.

2. The method according to claim 1, characterized in that Before spatial indexing based on truck stop data within a preset time period, it also includes: Perform stop analysis on the truck trajectory data within a preset time period to obtain the stop start point and stop end point; Calculate the stop duration based on the stop start point and the stop end point; The coordinates of the stop point are obtained based on the average values ​​of the horizontal coordinates and the vertical coordinates of the stop start point coordinates and the stop end point coordinates; The truck stop data is obtained based on the stop duration, stop point coordinates and vehicle identification.

3. The method according to claim 2, characterized in that After obtaining the truck stop data, the method further includes: Based on the matching of the stop point coordinates with the map data, obtaining the point of interest information corresponding to the stop point coordinates; Based on the information of points of interest corresponding to the coordinates of the stop points, filtering is performed to filter out the stop points in the public parking area that are not trucks, and to obtain filtered truck stop data.

4. The method according to claim 1, characterized in that: Based on the statistical analysis of H3 grid identification, we obtained the high-heat H3 grid, including: All truck stops are aggregated according to the H3 grid identification; The stop frequency and number of parked vehicles in each grid are counted based on the H3 grid identification; The H3 grids whose stop frequency is greater than a preset first threshold and whose number of parked vehicles is greater than a preset second threshold are screened out to obtain the high-heat H3 grid.

5. The method according to claim 1, characterized in that Based on the grid center coordinates corresponding to the high-heat H3 grid, remote sensing images of the docking area and the surrounding area are captured, including: Based on the center coordinates of the grid, a remote sensing image corresponding to the docking area is captured as a center remote sensing image; Taking the center coordinates of the grid as the center, obtain the remote sensing image pictures of the surrounding eight H3 grids as the surrounding remote sensing image map; The central remote sensing image and the peripheral remote sensing image are spliced ​​into a nine-square remote sensing image.

6. The method according to claim 1, characterized in that Before inputting the remote sensing image into the pre-trained truck parking area recognition model, the method further includes: Obtain multiple remote sensing images of the docking area and its surrounding areas, and annotate them to obtain a training set and a test set; Use the wide residual network model to perform multiple iterations of training on the training set, and generate a model file for each iteration; Evaluate the model files on the test set and calculate the precision and recall of each model file by comparing the model output with the actual labels; Based on the accuracy and recall rate, the model file with the best performance is selected as the trained truck parking area recognition model.

7. The method according to claim 1, characterized in that The remote sensing image is input into a pre-trained truck parking area recognition model, and the recognition result of the truck parking area is output, including: The remote sensing image is input into a pre-trained truck parking area recognition model, and the recognition result of the truck parking area, the corresponding H3 grid mark and the grid center coordinates are output.

8. A truck parking area identification device, characterized in that: include: The stop analysis module is used to perform spatial indexing based on the truck stop data within a preset time period to obtain the H3 grid corresponding to the truck stop point; The screening module is used to perform statistical analysis based on H3 grid identifiers to obtain high-heat H3 grids; A remote sensing image acquisition module is used to capture the remote sensing image of the docking area and the surrounding area based on the grid center coordinates corresponding to the high-heat H3 grid; The recognition module is used to input the remote sensing image into a pre-trained truck parking area recognition model and output a recognition result of the truck parking area.

9. An electronic device, characterized in that: The system comprises a processor and a memory storing program instructions, wherein the processor is configured to execute the method for identifying a truck parking area according to any one of claims 1 to 7 when executing the program instructions.

10. A computer-readable medium, characterized in that Computer-readable instructions are stored thereon, and the computer-readable instructions are executed by a processor to implement a method for identifying a truck parking area as described in any one of claims 1 to 7.

Images