Method and apparatus for determining location

Abstract

Embodiments of the present disclosure provide a position determination method and device. The method comprises: after receiving a target order corresponding to a user, obtaining a delivery position corresponding to the user; determining a target road section corresponding to the target order according to the delivery position; generating a trajectory data distribution image according to a trajectory point feature corresponding to a target trajectory point in the target road section and a target group address corresponding to the delivery position; inputting the trajectory data distribution image into a target position prediction model, and obtaining a target position corresponding to the target order output by the target position prediction model. Embodiments of the present disclosure can improve the accuracy of position prediction.

Description

Technical Field

[0001] The present disclosure relates to the field of location positioning technology, and in particular to a location determination method and apparatus. Background Technology

[0002] In the order delivery process, the final delivery stage is a crucial step that directly impacts the experience of both users and delivery personnel. The location of the POI (Point of Interest) coordinates at the destination address has a significant impact on the delivery personnel's navigation route, delivery fee pricing, and estimated delivery time.

[0003] Currently, there are two main methods for obtaining POIs: one is the master database data collected by frontline workers over a long period of time, which is widely used by various map providers; the other is the POI coordinate generation algorithm based on delivery personnel check-in data.

[0004] Algorithms based on delivery personnel check-in data largely solve the problems of high reliance on manpower and the inability to standardize clustering algorithm parameters. However, its drawback is its high dependence on the quality of the delivery personnel check-in data. Without a scoring mechanism, many bad cases will occur. Even with a scoring mechanism, although bad cases are filtered out, some delivery personnel check-in points are too scattered, making it impossible to calculate the correct coordinates. Summary of the Invention

[0005] The embodiments of this disclosure provide a location determination method and apparatus to improve the accuracy of location prediction.

[0006] According to a first aspect of the embodiments of this disclosure, a location determination method is provided, comprising:

[0007] After receiving the target order for the user, obtain the delivery location corresponding to the user;

[0008] Based on the delivery location, determine the target road segment corresponding to the target order;

[0009] Based on the trajectory point features corresponding to the target trajectory points within the target road segment and the target group address corresponding to the delivery location, a trajectory data distribution image is generated;

[0010] The trajectory data distribution image is input into the target location prediction model, and the target location corresponding to the target order is obtained from the target location prediction model.

[0011] Optionally, before obtaining the delivery location corresponding to the user, the method further includes:

[0012] Obtain the delivery trajectory of the delivery personnel associated with the historical orders, and the user location corresponding to the historical orders;

[0013] Based on the user's location and the delivery trajectory, determine the historical route segment corresponding to the historical order;

[0014] Based on the trajectory point features corresponding to the historical trajectory points within the historical road segment, a multi-source data distribution image is generated;

[0015] The multi-source data distribution image is input into the location prediction model to be trained, so as to obtain a target location prediction model associated with the user's location.

[0016] Optionally, determining the historical route segment corresponding to the historical order based on the user's location and the delivery trajectory includes:

[0017] Based on the user's location and the delivery trajectory, determine the delivery location point that matches the historical order;

[0018] The delivery location points are segmented to obtain multiple segmented delivery routes;

[0019] Based on the location characteristics of each location point in the route delivery path, the historical route segment corresponding to the historical order is determined.

[0020] Optionally, the segmentation of the delivery location to obtain multiple segmented delivery routes includes:

[0021] The delivery location points are denoised according to a preset distance threshold to obtain denoised processed delivery location points.

[0022] The processing and delivery locations are divided equally according to the preset number of locations to obtain multiple segmented delivery routes.

[0023] Optionally, determining the historical route segment corresponding to the historical order based on the location features of each location point in the branch delivery route includes:

[0024] Based on the corner features corresponding to the delivery location points, the quadrant value corresponding to each delivery location point is calculated;

[0025] Based on the processing and delivery location points within the segmented delivery routes and the limit values, the historical segments in the multiple segmented delivery routes are determined.

[0026] Optionally, generating a multi-source data distribution image based on the trajectory point features corresponding to historical trajectory points within the historical road segment includes:

[0027] Based on the address corresponding to the user's location, the user's location is grouped into address groups to obtain grouped addresses;

[0028] Generate a blank area image associated with the group address;

[0029] The historical trajectory points and the group addresses are filled into the blank area of ​​the image to generate the multi-source data distribution image corresponding to the group addresses.

[0030] Optionally, the step of inputting the multi-source data distribution image into the location prediction model to be trained, so as to train the location prediction model to obtain a target location prediction model associated with the user's location, includes:

[0031] Based on the address truth value associated with the group address, the multi-source data distribution image is labeled to generate a sample image corresponding to the group address;

[0032] The sample image is input into the location prediction model to be trained, so as to train the location prediction model to obtain a target location prediction model associated with the user's location.

[0033] According to a second aspect of the embodiments of the present disclosure, a position determination apparatus is provided, comprising:

[0034] The delivery location acquisition module is used to acquire the delivery location corresponding to the user after receiving the target order corresponding to the user;

[0035] The target route segment determination module is used to determine the target route segment corresponding to the target order based on the delivery location;

[0036] The distribution image generation module is used to generate a trajectory data distribution image based on the trajectory point features corresponding to the target trajectory points within the target road segment and the target group address corresponding to the delivery location;

[0037] The target location acquisition module is used to input the trajectory data distribution image into the target location prediction model and obtain the target location corresponding to the target order output by the target location prediction model.

[0038] Optionally, the device further includes:

[0039] The delivery trajectory acquisition module is used to acquire the delivery trajectory of the delivery personnel associated with historical orders, as well as the user location corresponding to the historical orders;

[0040] The historical route determination module is used to determine the historical route corresponding to the historical order based on the user's location and the delivery trajectory.

[0041] The multi-source image generation module is used to generate a multi-source data distribution image based on the trajectory point features corresponding to the historical trajectory points within the historical road segment.

[0042] The target model acquisition module is used to input the multi-source data distribution image into the location prediction model to be trained, so as to train the location prediction model to obtain a target location prediction model associated with the user's location.

[0043] Optionally, the historical road segment determination module includes:

[0044] The delivery location point determination unit is used to determine the delivery location point that matches the historical order based on the user's location and the delivery trajectory.

[0045] The segmented delivery route acquisition unit is used to segment the delivery location points to obtain multiple segmented delivery routes;

[0046] The historical route segment determination unit is used to determine the historical route segment corresponding to the historical order based on the location characteristics of each location point in the branch delivery route.

[0047] Optionally, the segmented delivery route acquisition unit includes:

[0048] The noise-reducing location point acquisition subunit is used to perform noise reduction processing on the delivery location point according to a preset distance threshold to obtain the noise-reduced processed delivery location point.

[0049] The segmented delivery route acquisition subunit is used to equally divide the processing delivery location points according to the preset number of location points to obtain multiple segmented delivery routes.

[0050] Optionally, the historical road segment determination unit includes:

[0051] The limit value calculation subunit is used to calculate the limit value corresponding to each delivery location point based on the corner features corresponding to the delivery location point;

[0052] The historical route segment determination subunit is used to determine the historical routes in the multiple segmented delivery routes based on the processing delivery location points within the segmented delivery routes and the limit values.

[0053] Optionally, the multi-source image generation module includes:

[0054] The group address acquisition unit is used to group the user's location according to the address corresponding to the user's location to obtain the group address;

[0055] The blank area image generation unit is used to generate a blank area image associated with the group address;

[0056] A multi-source distribution image generation unit is used to fill the blank area image with the historical trajectory points and the group address to generate the multi-source data distribution image corresponding to the group address.

[0057] Optionally, the target model acquisition module includes:

[0058] The sample image generation unit is used to perform labeling processing on the multi-source data distribution image based on the address truth value associated with the group address, and generate the sample image corresponding to the group address;

[0059] The target location model acquisition unit is used to input the sample image into the location prediction model to be trained, so as to train the location prediction model to obtain a target location prediction model associated with the user's location.

[0060] According to a third aspect of the embodiments of this disclosure, an electronic device is provided, comprising:

[0061] A processor, a memory, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the program, implements any of the preceding location determination methods.

[0062] According to a fourth aspect of the embodiments of the present disclosure, a readable storage medium is provided, which, when instructions in the storage medium are executed by a processor of an electronic device, enables the electronic device to perform the location determination method described in any of the preceding claims.

[0063] This disclosure provides a location determination method and apparatus. After receiving a target order from a user, the method obtains the user's delivery location. Based on the delivery location, it determines the target road segment corresponding to the target order. Using the trajectory point features of target trajectory points within the target road segment and the target group address corresponding to the delivery location, it generates a trajectory data distribution image. This image is then input into a target location prediction model, and the target location corresponding to the target order is obtained from the model's output. This disclosure combines multiple location data sources as input, not limited to delivery personnel check-in data, and processes the distribution image corresponding to user and delivery personnel location data. This transforms the geographic data mining problem into an image processing problem, improving the accuracy of location prediction. Attached Figure Description

[0064] To more clearly illustrate the technical solutions of the embodiments of this disclosure, the accompanying drawings used in the description of the embodiments of this disclosure will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the embodiments of this disclosure. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0065] Figure 1 A flowchart illustrating the steps of a location determination method provided for embodiments of this disclosure;

[0066] Figure 2 A flowchart illustrating the steps of another location determination method provided for embodiments of this disclosure;

[0067] Figure 3 A schematic diagram of a planar octagon provided for an embodiment of this disclosure;

[0068] Figure 4 A schematic diagram of a position determination device provided for an embodiment of this disclosure;

[0069] Figure 5 This is a schematic diagram of another position determination device provided for an embodiment of the present disclosure. Detailed Implementation

[0070] The technical solutions of the embodiments of this disclosure will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this disclosure. All other embodiments obtained by those skilled in the art based on the embodiments of this disclosure without creative effort are within the protection scope of the embodiments of this disclosure.

[0071] Example 1

[0072] Reference Figure 1 The diagram illustrates a flowchart of the steps of a location determination method provided by an embodiment of this disclosure, as shown below. Figure 1 As shown, the method for determining the location may specifically include the following steps:

[0073] Step 101: After receiving the target order corresponding to the user, obtain the delivery location corresponding to the user.

[0074] The embodiments disclosed herein can be applied to scenarios where user location is obtained from trajectory data distribution images determined by combining multi-source data.

[0075] A target order refers to an order selected by the user that needs to be delivered. In this example, a target order can be a food delivery order or a logistics order, etc. The type of target order can be determined according to the actual situation, and this embodiment does not impose any restrictions on it.

[0076] Delivery location refers to the location reserved by the user on the order platform when placing an order. For example, if the user reserves three locations on the order platform, namely location 1, location 2 and location 3, when the user selects to pay for the target order, they can choose one of the locations as the delivery location of the target order, such as choosing location 1 or location 3, etc.

[0077] Of course, in this embodiment, the delivery location can be the location where the order is delivered or the location where the order is picked up. For example, when the target order is an online ride-hailing order, the user will provide a pick-up location, which can be used as the delivery location.

[0078] It is understood that the above examples are merely examples listed to better understand the technical solutions of the embodiments of this disclosure, and are not intended to be the only limitation of this embodiment.

[0079] After receiving a user's target order, the corresponding delivery location can be obtained.

[0080] After obtaining the delivery location, proceed to step 102.

[0081] Step 102: Determine the target road segment corresponding to the target order based on the delivery location.

[0082] The target route segment refers to the delivery route segment associated with the target order. For example, when the target order is a food delivery order, the target route segment can be the delivery route segment; when the target order is a ride-hailing order, the target route segment can be the route segment where the vehicle picks up the user, etc.

[0083] After obtaining the user's delivery location, the target route for the target order can be determined based on the delivery location. Specifically, the delivery location can be found on the electronic map, and then the route associated with the map location can be found on the electronic map, which is the route that can be delivered to the delivery location.

[0084] Of course, in the actual implementation, the road segments associated with different locations can be saved in advance in the order system. After obtaining the user's corresponding delivery location, the target road segments associated with the delivery location can be directly obtained from the system database.

[0085] After determining the target road segment corresponding to the target order based on the delivery location, proceed to step 103.

[0086] Step 103: Generate a trajectory data distribution image based on the trajectory point features corresponding to the target trajectory points within the target road segment and the target group address corresponding to the delivery location.

[0087] A trajectory data distribution image refers to an image determined by combining the characteristics of trajectory points corresponding to target trajectory points within a target road segment.

[0088] After obtaining the target road segment, the target trajectory points within the target road segment can be obtained. The trajectory point features of the target trajectory points and the target group address corresponding to the delivery location are then filled into the blank area image associated with the delivery location. In this way, a trajectory data distribution image can be generated. Specifically, the positions of the target trajectory points and target group addresses in the real map can be drawn in the blank area image based on their positions in the real map, thereby obtaining the trajectory data distribution image.

[0089] The detailed process of obtaining the distribution image and target group address will be described in detail in the model training section of Embodiment 2 below, and will not be repeated here.

[0090] After generating a trajectory data distribution image based on the trajectory point characteristics corresponding to the target trajectory points within the target road segment and the target group address corresponding to the delivery location, step 104 is executed.

[0091] Step 104: Input the trajectory data distribution image into the target location prediction model, and obtain the target location corresponding to the target order output by the target location prediction model.

[0092] The target location prediction model refers to a model that is pre-trained using a distribution image to predict the user's accurate location. The training process of the target location prediction model will be described in detail in the following Example 2, and will not be repeated here.

[0093] The target location refers to the accurate delivery location of the target order output by the target location prediction model.

[0094] After generating the trajectory data distribution image, the trajectory data can be input into the target location prediction model to obtain the target location corresponding to the target order output by the target location prediction model.

[0095] This embodiment combines multiple location data as input data sources, not limited to delivery personnel check-in data, and processes the distribution images corresponding to user and delivery personnel location data, transforming the geographic data mining problem into an image processing problem, thereby improving the accuracy of location prediction.

[0096] The location determination method provided in this disclosure involves obtaining the user's delivery location after receiving a target order, determining the target road segment corresponding to the target order based on the delivery location, generating a trajectory data distribution image based on the trajectory point features of the target trajectory points within the target road segment and the target group address corresponding to the delivery location, inputting the trajectory data distribution image into a target location prediction model, and obtaining the target location corresponding to the target order output by the target location prediction model. This disclosure combines multiple location data as input data sources, not limited to delivery personnel check-in data, and processes the distribution image corresponding to the user and delivery personnel location data, transforming the geographic data mining problem into an image processing problem, thereby improving the accuracy of location prediction.

[0097] Example 2

[0098] Reference Figure 2 The diagram illustrates a flowchart of the steps of a location determination method provided by an embodiment of this disclosure, as shown below. Figure 2 As shown, the method for determining the location may specifically include the following steps:

[0099] Step 201: Obtain the delivery trajectory of the delivery personnel associated with the historical order, and the user location corresponding to the historical order.

[0100] The embodiments disclosed herein can be applied to scenarios where location prediction models are trained using multi-source data.

[0101] Historical orders refer to orders that have been delivered and are saved in the order system.

[0102] User location refers to the address provided by the user when placing an order. For example, when a user places a food delivery order, they will add the delivery location, which is the user's location. Similarly, when a user hails a ride online, they will provide a pick-up location, which is the user's location.

[0103] It is understood that the above examples are merely examples listed to better understand the technical solutions of the embodiments of this disclosure, and are not intended to be the only limitation of this embodiment.

[0104] When training a location prediction model is required, historical orders saved in the order system can be obtained, and based on the delivery trajectory of the delivery personnel associated with the historical orders and the user's location, step 202 can be executed.

[0105] Step 202: Based on the user's location and the delivery trajectory, determine the historical route segment corresponding to the historical order.

[0106] Historical route segments refer to delivery and / or pickup routes associated with historical orders. For example, when the historical order is a food delivery order, the target route segment may include both the pickup and delivery routes, while when the historical order is a ride-hailing order, the target route segment may be the route for picking up and dropping off the user.

[0107] It is understood that the above examples are merely examples listed to better understand the technical solutions of the embodiments of this disclosure, and are not intended to be the only limitation of this embodiment.

[0108] After obtaining the user's location and delivery trajectory associated with historical orders, the historical road segments corresponding to the historical orders can be determined based on the user's location and delivery trajectory. Specifically, historical road segments can be filtered out based on the characteristics of the trajectory points in the delivery trajectory. The specific implementation method described below can be described in detail.

[0109] In one specific implementation of this disclosure, step 202 may include:

[0110] Sub-step A1: Determine the delivery location point that matches the historical order based on the user's location and the delivery trajectory.

[0111] In this embodiment, the delivery location point refers to the user's location and the location point in the delivery trajectory.

[0112] Delivery trajectory refers to the path taken by delivery personnel during the process of completing historical orders. It includes the trajectory from the order acceptance location to the user's location. For example, when the historical order is a food delivery order, the delivery trajectory includes the entire process from the order acceptance location to the merchant to pick up the order, and then from the merchant to the user's location. The path in the entire process can be defined as the delivery trajectory.

[0113] After obtaining the user's location and the delivery trajectory of the delivery personnel associated with historical orders, the delivery location point that matches the historical order can be determined based on the user's location and the delivery trajectory.

[0114] After determining the delivery location that matches the historical order, proceed to sub-step A2.

[0115] Sub-step A2: Segment the delivery location points to obtain multiple segmented delivery routes.

[0116] Segmented delivery routes refer to multiple segmented routes obtained after dividing the delivery location points into segments.

[0117] After obtaining the delivery location points that match historical orders, these delivery location points can be segmented to obtain multiple segmented delivery routes. Specifically, this can be described in detail below with reference to the specific implementation methods.

[0118] In another specific implementation of this disclosure, the above sub-step A2 may include:

[0119] Sub-step B1: Denoise the delivery location points according to a preset distance threshold to obtain the denoised delivery location points.

[0120] In this embodiment, after obtaining the delivery location points, it is also necessary to remove abnormal location points, i.e., to perform noise reduction processing. Specifically, the delivery location points can be denoised using a preset distance threshold to obtain denoised delivery location points. In this embodiment, due to partial data loss during acquisition, many noise points appear in the middle, which are far away from other trajectory points. Therefore, the second step is to remove these noise points using a distance threshold. The distance threshold can be set to M meters (M is a positive number), meaning that a point is discarded if the distance between it and the points before and after it exceeds M meters.

[0121] After denoising the delivery location points according to the preset distance threshold to obtain the denoised delivery location points, sub-step B2 is executed.

[0122] Sub-step B2: Divide the processing and delivery location points equally according to the preset number of location points to obtain multiple segmented delivery paths.

[0123] After obtaining the denoised delivery location points, the delivery location points can be divided equally according to the preset number of location points to obtain multiple segmented delivery paths. Specifically, a sliding window can be preset with a size of K (K is a positive integer) location points. Then, the delivery location points are covered segment by segment through the sliding window. At this time, each K location point can form a segmented delivery path.

[0124] After segmenting the delivery location points to obtain multiple segmented delivery routes, sub-step A3 is executed.

[0125] Sub-step A3: Based on the location characteristics of each location point in the branch delivery path, determine the historical route segment corresponding to the historical order.

[0126] After segmenting the delivery location points to obtain multiple segmented delivery routes, the historical route corresponding to the historical order can be determined based on the location point characteristics of each location point in the segmented delivery route. Specifically, the following detailed implementation method can be described in detail.

[0127] In another specific implementation of this disclosure, sub-step A3 may include:

[0128] Sub-step C1: Calculate the quagmire value corresponding to each delivery location point based on the corner features corresponding to the delivery location point.

[0129] In this embodiment, the limit value refers to the segmented limit value of the plane angle used to indicate the delivery location point.

[0130] After obtaining the delivery location points, the corresponding zoning values ​​for each delivery location point can be calculated based on the corner features corresponding to those points, such as... Figure 3 As shown, after obtaining the delivery location, the corner features corresponding to the delivery location can be obtained based on the GPS positioning system. Then, the corresponding zodiac value can be calculated by combining the corner features. A zodiac interval can be defined as 36 degrees. Figure 3 As shown.

[0131] After calculating the limit value corresponding to each delivery location point based on the corner characteristics corresponding to the delivery location point, sub-step C2 is executed.

[0132] Sub-step C2: Based on the processing delivery location points within the segmented delivery routes and the limit values, determine the historical segments in the multiple segmented delivery routes.

[0133] After obtaining the segmented delivery routes, historical segments corresponding to historical orders can be determined based on the location characteristics (i.e., corner features, stop points, and speed features) of each location point within the segmented delivery routes. This involves identifying historical segments across multiple segmented delivery routes based on the processing delivery location points and boundary values ​​within the segmented delivery routes. For example, regarding the corner feature, in delivery and pickup segments, ten consecutive points typically cover more than N boundary values; that is, ten corners exceed a preset angle and include U-turns. Therefore, a corner exceeding N boundary values ​​can be considered a potential pickup or delivery segment. A stop point is the duration a rider remains within a preset range at the first trajectory point where their speed is less than Hm / s. Riders only stop when encountering red lights, traffic accidents, or delivery / pickup scenarios. Since red light or traffic accident scenarios do not involve significant corner features, the combination of corners and stops greatly increases the reliability of identifying delivery / pickup scenarios. The trajectory speed is actually already included in the rest point, because riders cannot maintain a high speed during delivery / pickup scenarios. Therefore, a speed of less than H / s is a necessary condition for determining the delivery / pickup scenario.

[0134] After determining the historical route segment corresponding to the historical order based on the user's location and delivery trajectory, proceed to step 203.

[0135] Step 203: Generate a multi-source data distribution image based on the trajectory point features corresponding to the historical trajectory points within the historical road segment.

[0136] Multi-source data distribution images refer to images generated by combining user location and historical trajectory points within historical road segments in the delivery trajectory.

[0137] After identifying historical road segments, a multi-source data distribution image can be generated based on the trajectory point features corresponding to historical trajectory points within those segments. Specifically, the image can be annotated with three dimensions of features—turning angles, stopping points, and speeds—corresponding to historical trajectory points, thus obtaining the multi-source data distribution image. The process of generating the multi-source data distribution image can be described in detail below with reference to the specific implementation method.

[0138] In another specific implementation of this disclosure, step 204 may include:

[0139] Sub-step D1: Based on the address corresponding to the user's location, group the user's location into address groups to obtain grouped addresses.

[0140] In this embodiment, the grouped address refers to the address obtained by grouping the user's location address.

[0141] After obtaining the user's location, an NLP parsing algorithm can be used to analyze the location. A multi-classification model can then parse the obtained user address to obtain the core keywords at different levels within the address. For example, for a residential community, the three dimensions needed are "community," "building number," and "unit number." The parsing results are shown in Table 1 below.

[0142] Table 1:

[0143]

[0144] The parsed result or the unparsed complete address is used as the name of the POI. The detailed address, building number, and unit number in Table 1 above are the three dimensions of the group address.

[0145] After grouping the user's location into address groups based on the address corresponding to the user's location, sub-step D2 is executed.

[0146] Sub-step D2: Generate a blank region image associated with the group address.

[0147] Blank area images refer to images generated without filled location information associated with packet addresses.

[0148] After grouping the user's location according to the address corresponding to the user's location to obtain the group address, a blank area image associated with the group address can be generated. Specifically, a blank image can be generated first, and then the group address can be filled into the blank image to obtain the blank area image associated with the group address.

[0149] After generating the blank region image associated with the group address, substep D3 is executed.

[0150] Sub-step D3: Fill the blank area image with the historical trajectory points and the group address to generate the multi-source data distribution image corresponding to the group address.

[0151] After generating a blank area image associated with the group address, historical trajectory points and group addresses can be filled into the blank area image, thereby generating a multi-source data distribution image corresponding to the group address. Specifically, based on the positions of historical trajectory points and group addresses in the real map, the positions of historical trajectory points and group addresses in the real map can be drawn in the blank area image to obtain a multi-source data distribution image corresponding to the group address.

[0152] After obtaining the multi-source data distribution image, proceed to step 204.

[0153] Step 204: Input the multi-source data distribution image into the location prediction model to be trained, so as to train the location prediction model to obtain a target location prediction model associated with the user's location.

[0154] After generating a multi-source data distribution image, the multi-source data distribution image can be input into the location prediction model to be trained, so as to obtain a target location prediction model associated with the user's location.

[0155] Furthermore, in the subsequent process, the target location prediction model can be used to accurately predict the delivery location of the order by combining the user's order placement location (i.e., the user's location).

[0156] The process of obtaining the target location prediction model described above can be described in detail in conjunction with the following specific implementation methods.

[0157] In another specific implementation of this disclosure, step 204 may include:

[0158] Sub-step E1: Based on the address truth value associated with the group address, label the multi-source data distribution image to generate a sample image corresponding to the group address.

[0159] In this embodiment, after generating the multi-source data distribution image, the ground truth values ​​associated with the group addresses can be obtained. The multi-source data distribution image is then labeled based on these ground truth values ​​to generate sample images corresponding to the group addresses. Specifically, a convolutional neural network is used to predict the true location of the POIs. First, some ground truth values ​​are selected from the parent database using a combination of manual visualization and rule-based methods to label the samples. The label is a matrix of the same size as the geohash box. Ground truth positions are 1, and other positions are 0.

[0160] After labeling the multi-source data distribution image based on the address truth value associated with the group address, sample images corresponding to the group address can be generated.

[0161] After generating the sample image corresponding to the group address, execute sub-step E2.

[0162] Sub-step E2: Input the sample image into the location prediction model to be trained, so as to train the location prediction model to obtain a target location prediction model associated with the user's location.

[0163] After generating sample images corresponding to the grouped addresses, these images can be input into the location prediction model to be trained, resulting in a target location prediction model associated with the user's location. Specifically, the input data consists of three data distribution images and a label matrix of the same size. The training model can be a U-Net model or other convolutional neural network models. The final model will predict the geohash block with the highest probability as the POI coordinate location.

[0164] The embodiments of this disclosure combine multiple location data as input data sources, not limited to delivery personnel check-in data, and process based on the distribution images corresponding to user and delivery personnel location data, transforming the geographic data mining problem into an image processing problem, thereby improving the accuracy of location prediction.

[0165] Step 205: After receiving the target order corresponding to the user, obtain the delivery location corresponding to the user.

[0166] A target order refers to an order selected by the user that needs to be delivered. In this example, a target order can be a food delivery order or a logistics order, etc. The type of target order can be determined according to the actual situation, and this embodiment does not impose any restrictions on it.

[0167] Delivery location refers to the location reserved by the user on the order platform when placing an order. For example, if the user reserves three locations on the order platform, namely location 1, location 2 and location 3, when the user selects to pay for the target order, they can choose one of the locations as the delivery location of the target order, such as choosing location 1 or location 3, etc.

[0168] Of course, in this embodiment, the delivery location can be the location where the order is delivered or the location where the order is picked up. For example, when the target order is an online ride-hailing order, the user will provide a pick-up location, which can be used as the delivery location.

[0169] It is understood that the above examples are merely examples listed to better understand the technical solutions of the embodiments of this disclosure, and are not intended to be the only limitation of this embodiment.

[0170] After receiving a user's target order, the corresponding delivery location can be obtained.

[0171] After obtaining the delivery location, proceed to step 206.

[0172] Step 206: Determine the target road segment corresponding to the target order based on the delivery location.

[0173] The target route segment refers to the delivery route segment associated with the target order. For example, when the target order is a food delivery order, the target route segment can be the delivery route segment; when the target order is a ride-hailing order, the target route segment can be the route segment where the vehicle picks up the user, etc.

[0174] After obtaining the user's delivery location, the target road segment corresponding to the target order can be determined based on the delivery location. The process of obtaining the target road segment can refer to the above-mentioned process of obtaining the historical road segment, and will not be repeated here in this embodiment.

[0175] After determining the target road segment corresponding to the target order based on the delivery location, proceed to step 206.

[0176] Step 207: Generate a trajectory data distribution image based on the trajectory point features corresponding to the target trajectory points within the target road segment and the target group address corresponding to the delivery location.

[0177] A trajectory data distribution image refers to an image determined by combining the characteristics of trajectory points corresponding to target trajectory points within a target road segment.

[0178] After obtaining the target road segment, the target trajectory points within the target road segment can be obtained. The trajectory point features of the target trajectory points and the target group address corresponding to the delivery location are then filled into the blank area image associated with the delivery location. In this way, a trajectory data distribution image can be generated. Specifically, the positions of the target trajectory points and target group addresses in the real map can be drawn in the blank area image based on their positions in the real map, thereby obtaining the trajectory data distribution image.

[0179] The process of acquiring trajectory data distribution images can refer to the above-described process of acquiring multi-source data distribution images, and will not be repeated here in this embodiment.

[0180] After generating a trajectory data distribution image based on the trajectory point characteristics corresponding to the target trajectory points within the target road segment and the target group address corresponding to the delivery location, step 208 is executed.

[0181] Step 208: Input the trajectory data distribution image into the target location prediction model, and obtain the target location corresponding to the target order output by the target location prediction model.

[0182] The target location refers to the accurate delivery location of the target order output by the target location prediction model.

[0183] After generating the trajectory data distribution image, the trajectory data can be input into the target location prediction model to obtain the target location corresponding to the target order output by the target location prediction model.

[0184] This embodiment combines multiple location data as input data sources, not limited to delivery personnel check-in data, and processes the distribution images corresponding to user and delivery personnel location data, transforming the geographic data mining problem into an image processing problem, thereby improving the accuracy of location prediction.

[0185] The location determination method provided in this disclosure involves obtaining the user's delivery location after receiving a target order, determining the target road segment corresponding to the target order based on the delivery location, generating a trajectory data distribution image based on the trajectory point features of the target trajectory points within the target road segment and the target group address corresponding to the delivery location, inputting the trajectory data distribution image into a target location prediction model, and obtaining the target location corresponding to the target order output by the target location prediction model. This disclosure combines multiple location data as input data sources, not limited to delivery personnel check-in data, and processes the distribution image corresponding to the user and delivery personnel location data, transforming the geographic data mining problem into an image processing problem, thereby improving the accuracy of location prediction.

[0186] Example 3

[0187] Reference Figure 4 The diagram shows a structural schematic of a position determination device provided in an embodiment of this disclosure, as shown below. Figure 4 As shown, the position determination device 300 may specifically include the following modules:

[0188] The delivery location acquisition module 310 is used to acquire the delivery location corresponding to the user after receiving the target order corresponding to the user;

[0189] The target route determination module 320 is used to determine the target route corresponding to the target order based on the delivery location;

[0190] The distribution image generation module 330 is used to generate a trajectory data distribution image based on the trajectory point features corresponding to the target trajectory points within the target road segment and the target group address corresponding to the delivery location.

[0191] The target location acquisition module 340 is used to input the trajectory data distribution image into the target location prediction model and acquire the target location corresponding to the target order output by the target location prediction model.

[0192] The location determination apparatus provided in the embodiments of this disclosure obtains the user's delivery location after receiving a target order, determines the target road segment corresponding to the target order based on the delivery location, generates a trajectory data distribution image based on the trajectory point features corresponding to the target trajectory points within the target road segment and the target group address corresponding to the delivery location, inputs the trajectory data distribution image into a target location prediction model, and obtains the target location corresponding to the target order output by the target location prediction model. The embodiments of this disclosure combine multiple location data as input data sources, not limited to delivery personnel check-in data, and process the distribution image corresponding to the user and delivery personnel location data, transforming the geographic data mining problem into an image processing problem, thereby improving the accuracy of location prediction.

[0193] Example 4

[0194] Reference Figure 5 The diagram shows a structural schematic of another position determination device provided by an embodiment of the present disclosure, such as... Figure 5 As shown, the position determination device 400 may specifically include the following modules:

[0195] The delivery trajectory acquisition module 410 is used to acquire the delivery trajectory of the delivery personnel associated with the historical orders, and the user location corresponding to the historical orders;

[0196] The historical route determination module 420 is used to determine the historical route corresponding to the historical order based on the user's location and the delivery trajectory;

[0197] The multi-source image generation module 430 is used to generate a multi-source data distribution image based on the trajectory point features corresponding to the historical trajectory points within the historical road segment.

[0198] The target model acquisition module 440 is used to input the multi-source data distribution image into the location prediction model to be trained, so as to train the location prediction model to obtain a target location prediction model associated with the user's location.

[0199] The delivery location acquisition module 450 is used to acquire the delivery location corresponding to the user after receiving the target order corresponding to the user;

[0200] The target route determination module 460 is used to determine the target route corresponding to the target order based on the delivery location;

[0201] The distribution image generation module 470 is used to generate a trajectory data distribution image based on the trajectory point features corresponding to the target trajectory points within the target road segment and the target group address corresponding to the delivery location.

[0202] The target location acquisition module 480 is used to input the trajectory data distribution image into the target location prediction model and acquire the target location corresponding to the target order output by the target location prediction model.

[0203] Optionally, the historical road segment determination module 420 includes:

[0204] The delivery location point determination unit is used to determine the delivery location point that matches the historical order based on the user's location and the delivery trajectory.

[0205] The segmented delivery route acquisition unit is used to segment the delivery location points to obtain multiple segmented delivery routes;

[0206] The historical route segment determination unit is used to determine the historical route segment corresponding to the historical order based on the location characteristics of each location point in the branch delivery route.

[0207] Optionally, the segmented delivery route acquisition unit includes:

[0208] The noise-reducing location point acquisition subunit is used to perform noise reduction processing on the delivery location point according to a preset distance threshold to obtain the noise-reduced processed delivery location point.

[0209] The segmented delivery route acquisition subunit is used to equally divide the processing delivery location points according to the preset number of location points to obtain multiple segmented delivery routes.

[0210] Optionally, the historical road segment determination unit includes:

[0211] The limit value calculation subunit is used to calculate the limit value corresponding to each delivery location point based on the corner features corresponding to the delivery location point;

[0212] The historical route segment determination subunit is used to determine the historical routes in the multiple segmented delivery routes based on the processing delivery location points within the segmented delivery routes and the limit values.

[0213] Optionally, the multi-source image generation module 430 includes:

[0214] The group address acquisition unit is used to group the user's location according to the address corresponding to the user's location to obtain the group address;

[0215] The blank area image generation unit is used to generate a blank area image associated with the group address;

[0216] A multi-source distribution image generation unit is used to fill the blank area image with the historical trajectory points and the group address to generate the multi-source data distribution image corresponding to the group address.

[0217] Optionally, the target model acquisition module 440 includes:

[0218] The sample image generation unit is used to perform labeling processing on the multi-source data distribution image based on the address truth value associated with the group address, and generate the sample image corresponding to the group address;

[0219] The target location model acquisition unit is used to input the sample image into the location prediction model to be trained, so as to train the location prediction model to obtain a target location prediction model associated with the user's location.

[0220] The location determination apparatus provided in the embodiments of this disclosure obtains the user's delivery location after receiving a target order, determines the target road segment corresponding to the target order based on the delivery location, generates a trajectory data distribution image based on the trajectory point features corresponding to the target trajectory points within the target road segment and the target group address corresponding to the delivery location, inputs the trajectory data distribution image into a target location prediction model, and obtains the target location corresponding to the target order output by the target location prediction model. The embodiments of this disclosure combine multiple location data as input data sources, not limited to delivery personnel check-in data, and process the distribution image corresponding to the user and delivery personnel location data, transforming the geographic data mining problem into an image processing problem, thereby improving the accuracy of location prediction.

[0221] Embodiments of this disclosure also provide an electronic device, including: a processor, a memory, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the position determination method of the foregoing embodiments.

[0222] Embodiments of this disclosure also provide a readable storage medium that, when instructions in the storage medium are executed by a processor of an electronic device, enables the electronic device to perform the location determination method of the foregoing embodiments.

[0223] As the device embodiment is basically similar to the method embodiment, the description is relatively simple, and relevant parts can be found in the description of the method embodiment.

[0224] The algorithms and displays provided herein are not inherently related to any particular computer, virtual system, or other device. Various general-purpose systems can also be used in conjunction with the teachings herein. The required structure for constructing such systems is apparent from the above description. Furthermore, the embodiments of this disclosure are not directed to any particular programming language. It should be understood that the embodiments of this disclosure described herein can be implemented using various programming languages, and the above description of specific languages ​​is for the purpose of disclosing the best mode of implementation of the embodiments of this disclosure.

[0225] Numerous specific details are set forth in the specification provided herein. However, it will be understood that embodiments of this disclosure may be practiced without these specific details. In some instances, well-known methods, structures, and techniques have not been shown in detail so as not to obscure the understanding of this specification.

[0226] Similarly, it should be understood that, in order to simplify this disclosure and aid in understanding one or more of the various inventive aspects, in the foregoing description of exemplary embodiments of this disclosure, various features of embodiments of this disclosure are sometimes grouped together in a single embodiment, figure, or description thereof. However, this approach to disclosure should not be construed as reflecting an intention that the claimed embodiments of this disclosure require more features than expressly recited in each claim. Rather, as reflected in the following claims, inventive aspects lie in fewer than all features of a single foregoing disclosed embodiment. Therefore, the claims following the detailed description are hereby expressly incorporated into that detailed description, wherein each claim itself is a separate embodiment of the present disclosure.

[0227] Those skilled in the art will understand that modules in the device of the embodiments can be adaptively changed and placed in one or more devices different from that embodiment. Modules, units, or components in the embodiments can be combined into a single module, unit, or component, and further, they can be divided into multiple sub-modules, sub-units, or sub-components. Except where at least some of such features and / or processes or units are mutually exclusive, any combination can be used to combine all features disclosed in this specification (including the accompanying claims, abstract, and drawings) and all processes or units of any method or device so disclosed. Unless expressly stated otherwise, each feature disclosed in this specification (including the accompanying claims, abstract, and drawings) may be replaced by an alternative feature serving the same, equivalent, or similar purpose.

[0228] The various component embodiments of this disclosure can be implemented in hardware, or as software modules running on one or more processors, or a combination thereof. Those skilled in the art will understand that microprocessors or digital signal processors (DSPs) can be used in practice to implement some or all of the functions of some or all of the components in the motion picture generation device according to the embodiments of this disclosure. Embodiments of this disclosure can also be implemented as device or apparatus programs for performing some or all of the methods described herein. Such programs implementing embodiments of this disclosure can be stored on a computer-readable medium or can take the form of one or more signals. Such signals can be downloaded from an Internet website, provided on a carrier signal, or provided in any other form.

[0229] It should be noted that the above embodiments are illustrative of embodiments of this disclosure and not restrictive of embodiments of this disclosure, and that those skilled in the art can devise alternative embodiments without departing from the scope of the appended claims. In the claims, any reference signs placed between parentheses should not be construed as limiting the claims. The word "comprising" does not exclude the presence of elements or steps not listed in a claim. The word "a" or "an" preceding an element does not exclude the presence of a plurality of such elements. Embodiments of this disclosure can be implemented by means of hardware comprising several different elements and by means of a suitably programmed computer. In the unit claims enumerating several means, several of these means may be embodied by the same item of hardware. The use of the words first, second, and third, etc., does not indicate any order. These words can be interpreted as names.

[0230] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0231] The above description is only a preferred embodiment of the present disclosure and is not intended to limit the embodiments of the present disclosure. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the embodiments of the present disclosure should be included within the protection scope of the embodiments of the present disclosure.

[0232] The above description is merely a specific implementation of the embodiments of this disclosure, but the protection scope of the embodiments of this disclosure is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the embodiments of this disclosure should be included within the protection scope of the embodiments of this disclosure. Therefore, the protection scope of the embodiments of this disclosure should be determined by the protection scope of the claims.

Claims

1. A method of position determination, characterized by include: After receiving the target order for the user, obtain the delivery location corresponding to the user; Based on the delivery location, the target route corresponding to the target order is determined, where the target route refers to the delivery route associated with the target order; Based on the trajectory point features corresponding to the target trajectory points within the target road segment and the location of the target group address corresponding to the delivery location on the real map, the locations of the target trajectory points and the target group address corresponding to the delivery location on the real map are plotted in the blank area image to generate a trajectory data distribution image; the trajectory data distribution image is input into the target location prediction model, and the target location corresponding to the target order output by the target location prediction model is obtained, wherein the target location refers to the accurate delivery location of the target order output by the target location prediction model.

2. The method of claim 1, wherein, Before obtaining the delivery location corresponding to the user, the method further includes: Obtain the delivery trajectory of the delivery personnel associated with the historical orders, and the user location corresponding to the historical orders; Based on the user's location and the delivery trajectory, determine the historical route segment corresponding to the historical order; Based on the trajectory point features corresponding to the historical trajectory points within the historical road segment, a multi-source data distribution image is generated; The multi-source data distribution image is input into the location prediction model to be trained, so as to obtain a target location prediction model associated with the user's location.

3. The method of claim 2, wherein, The step of determining the historical route segment corresponding to the historical order based on the user's location and the delivery trajectory includes: Based on the user's location and the delivery trajectory, determine the delivery location point that matches the historical order; The delivery location points are segmented to obtain multiple segmented delivery routes; Based on the location characteristics of each location point in the segmented delivery route, the historical route segment corresponding to the historical order is determined.

4. The method of claim 3, wherein, The process of segmenting the delivery location points to obtain multiple segmented delivery routes includes: The delivery location points are denoised according to a preset distance threshold to obtain denoised processed delivery location points. The processing and delivery locations are divided equally according to the preset number of locations to obtain multiple segmented delivery routes.

5. The method of claim 4, wherein, The step of determining the historical route segment corresponding to the historical order based on the location characteristics of each location point in the segmented delivery route includes: Based on the corner features corresponding to the delivery location points, the quadrant value corresponding to each delivery location point is calculated; Based on the processing and delivery location points within the segmented delivery routes and the limit values, the historical segments in the multiple segmented delivery routes are determined.

6. The method of claim 2, wherein, The step of generating a multi-source data distribution image based on the trajectory point features corresponding to historical trajectory points within the historical road segment includes: Based on the address corresponding to the user's location, the user's location is grouped into address groups to obtain grouped addresses; Generate a blank area image associated with the group address; The historical trajectory points and the group addresses are filled into the blank area of ​​the image to generate the multi-source data distribution image corresponding to the group addresses.

7. The method of claim 6, wherein, The step of inputting the multi-source data distribution image into the location prediction model to be trained, so as to train the location prediction model to obtain a target location prediction model associated with the user's location, includes: Based on the address truth value associated with the group address, the multi-source data distribution image is labeled to generate a sample image corresponding to the group address; The sample image is input into the location prediction model to be trained, so as to train the location prediction model to obtain a target location prediction model associated with the user's location.

8. A position determining apparatus, characterized by include: The delivery location acquisition module is used to acquire the delivery location corresponding to the user after receiving the target order corresponding to the user; The target route segment determination module is used to determine the target route segment corresponding to the target order based on the delivery location, wherein the target route segment refers to the delivery route segment associated with the target order; The distribution image generation module is used to draw the positions of the target trajectory points and the target group addresses corresponding to the delivery locations on the real map within a blank area image based on the trajectory point features corresponding to the target trajectory points within the target road segment and the positions of the target group addresses corresponding to the delivery locations on the real map, thereby generating a trajectory data distribution image. The target location acquisition module is used to input the trajectory data distribution image into the target location prediction model and obtain the target location corresponding to the target order output by the target location prediction model. The target location refers to the accurate delivery location of the target order output by the target location prediction model.

9. An electronic device, comprising: include: A processor, a memory, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the program, implements the location determination method according to any one of claims 1 to 7.

10. A readable storage medium, characterized by, When the instructions in the storage medium are executed by the processor of the electronic device, the electronic device is able to perform the location determination method according to any one of claims 1 to 7.

Images