Data processing method and related device
By using data processing methods in off-road activities, the off-road road book and rescue routes are adjusted in real time, the problem of low intelligence in traditional road book processing emergencies is solved, and more efficient obstacle response and rescue services are achieved.
Patent Information
- Application Number
- CN202411212550.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-30
- Publication Date
- 2024-11-08
AI Technical Summary
Traditional off-road road books rely on manual experience to generate, and the degree of intelligence of handling emergencies is relatively low, especially when encountering mudslides, landslides, vehicle failures, injuries, etc., lack of intelligent processing capabilities.
Using data processing method, the initial road book is generated by obtaining a three-dimensional map based on parameter configuration information, and dynamically adjusting the road book when the forward obstacle information is detected in real time to generate the first target road book; when receiving the rescue request, it is determined based on the current data to be rescued. Target rescue route.
It has improved the intelligent adjustment capabilities of Lushu, can timely match obstacles, ensure safe driving, and provide rapid rescue in emergencies to avoid casualties.
Smart Images

Figure CN118918282A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the intersection of intelligent transportation, big data processing and artificial intelligence, and more specifically, to a data processing method and related devices. Background Art
[0002] With the popularity of off-road activities and the increase in demand for outdoor adventures, traditional off-road road books often rely on manual experience. When performing off-road activities according to the road book, if an emergency (such as mudslides, landslides, vehicle failures, injuries, etc.) is encountered, it is generally handled manually based on experience, and the degree of intelligence in handling emergencies is low. Summary of the invention
[0003] In view of this, the present invention provides a data processing method and related devices to solve the problem of low intelligence level in processing emergencies.
[0004] In order to solve the above technical problems, the present invention adopts the following technical solutions:
[0005] A data processing method, comprising:
[0006] Obtain an initial road book generated based on parameter configuration information and a pre-configured three-dimensional map;
[0007] In the process of performing activities according to the initial road book, based on the detected front obstacle information, the initial road book is dynamically adjusted to obtain a first target road book, so as to perform activities based on the first target road book;
[0008] If a rescue request is received, a target rescue route is determined based on the active current rescue data, so as to perform a rescue operation based on the target rescue route.
[0009] Optionally, the three-dimensional map generation process includes:
[0010] Collect environmental data of the target area;
[0011] Using the environmental data, generating an initial three-dimensional map;
[0012] Each target is marked in the initial three-dimensional map to obtain a three-dimensional map.
[0013] Optionally, marking each target in the initial three-dimensional map to obtain a three-dimensional map includes:
[0014] Performing a feature extraction operation on the environmental data to obtain target features;
[0015] Using the target features, performing a semantic segmentation operation on the environmental data to identify various targets in the environmental data;
[0016] The various targets are marked in the initial three-dimensional map to obtain a three-dimensional map.
[0017] Optionally, obtaining an initial road book generated based on the parameter configuration information and a pre-configured three-dimensional map includes:
[0018] Obtain parameter configuration information, the parameter configuration information including: a starting point, an end point, and preference information;
[0019] Based on the three-dimensional map, a reference road book conforming to the parameter configuration information is generated using a large model;
[0020] The best road book is selected from the reference road book to obtain an initial road book.
[0021] Optionally, based on the detected front obstacle information, dynamically adjusting the initial road book to obtain a first target road book includes:
[0022] Obtaining front obstacle information detected by the drone; the front obstacle information includes obstacle information and / or severe weather information;
[0023] The large model is used to adjust the content of the initial road book based on the front obstacle information to obtain a first target road book.
[0024] Optionally, based on the current rescue data of the activity, determining the target rescue route includes:
[0025] Obtain the current rescue data collected by the drone;
[0026] Based on the current rescue data, planning a reference rescue route;
[0027] Analyzing the rescue risk of the reference rescue route;
[0028] A target rescue route whose rescue risk meets a preset rescue rule is selected from the reference rescue routes.
[0029] Optionally, in the process of carrying out activities according to the initial road book, the method further includes:
[0030] Obtain user adjustment demand information;
[0031] Based on the user adjustment demand information, the initial road book is dynamically adjusted to obtain a second target road book, so as to carry out activities based on the second target road book.
[0032] Optionally, in the process of carrying out activities according to the initial road book, the method further includes:
[0033] Detect the location information of the current activity;
[0034] determining navigation information based on the location information;
[0035] Output voice prompt information, where the voice prompt information includes the location information and the navigation information.
[0036] A data processing device, comprising:
[0037] A road book acquisition module is used to acquire an initial road book generated based on parameter configuration information and a pre-configured three-dimensional map;
[0038] A road book adjustment module is used to dynamically adjust the initial road book based on the detected front obstacle information during the process of performing activities according to the initial road book to obtain a first target road book, so as to perform activities based on the first target road book;
[0039] The rescue control module is used to determine a target rescue route based on the current rescue data of the activity if a rescue request is received, so as to perform a rescue operation based on the target rescue route.
[0040] An electronic device comprises at least one processor and a memory connected to the processor, wherein:
[0041] The memory is used to store computer programs;
[0042] The processor is used to execute the computer program so that the electronic device can implement the above-mentioned data processing method.
[0043] Compared with the prior art, the present invention has the following beneficial effects:
[0044] The present invention provides a data processing method and related devices. In the present invention, if the obstacle information ahead is detected, the initial road book is dynamically adjusted to obtain a first target road book, that is, the present invention can adjust the road book in real time based on the obstacle information, so that the adjusted road book matches the current obstacle situation, which is convenient for users to carry out activities according to the adjusted road book, and the road book adjustment has a high degree of intelligence. In addition, in the present invention, when a rescue request is received, the target rescue route can be determined based on the current rescue data of the activity, so as to perform rescue operations based on the target rescue route, and rescue services can be provided in time to avoid casualties. As can be seen from the above, the present invention provides handling measures when encountering obstacles and when rescue is needed, which can improve the degree of intelligence in handling emergencies. BRIEF DESCRIPTION OF THE DRAWINGS
[0045] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.
[0046] Figure 1 A flowchart of an off-road data processing method provided by an embodiment of the present invention;
[0047] Figure 2 A flowchart of a method for generating a three-dimensional off-road map provided by an embodiment of the present invention;
[0048] Figure 3 A flowchart of a method for marking an off-road map provided by an embodiment of the present invention;
[0049] Figure 4 A flow chart of a method for generating an off-road road book provided by an embodiment of the present invention;
[0050] Figure 5 A flowchart of a route planning method provided by an embodiment of the present invention;
[0051] Figure 6 A schematic structural diagram of an off-road data processing device provided by an embodiment of the present invention;
[0052] Figure 7 A schematic diagram of the structure of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION
[0053] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0054] With the popularity of off-road activities and the increase in demand for outdoor adventures, traditional off-road road books often rely on manual experience. When performing off-road activities according to the road book, if an emergency (such as mudslides, landslides, vehicle failures, injuries, etc.) is encountered, it is generally handled manually based on experience, and the degree of intelligence in handling emergencies is low.
[0055] In order to improve the intelligence of handling emergencies, the present invention provides a data processing method and related devices. In the present invention, if the obstacle information ahead is detected, the initial road book is dynamically adjusted to obtain a first target road book, that is, the present invention can adjust the road book in real time based on the obstacle information, so that the adjusted road book matches the current obstacle situation, which is convenient for users to carry out activities according to the adjusted road book, and the road book adjustment has a high degree of intelligence. In addition, in the present invention, when a rescue request is received, the target rescue route can be determined based on the current rescue data of the activity, so as to perform rescue operations based on the target rescue route, and rescue services can be provided in time to avoid casualties. From the above, it can be seen that the present invention provides processing measures when encountering obstacles and when rescue is needed, which can improve the intelligence of handling emergencies.
[0056] Based on the above content, an embodiment of the present invention provides a data processing method, and the execution subject can be a data processing device such as a vehicle controller, a server, a data processing center, etc. In the present invention, when a user is performing an activity (such as an off-road activity), he will carry a drone to explore the environment ahead. Therefore, the data processing device in this embodiment can interact with the drone in real time. If the user needs to request rescue, a rescue request can be sent to the data processing device through the drone, so that the data processing device can analyze the rescue route in time, inform the corresponding rescue personnel, and perform rescue operations in time to avoid casualties.
[0057] Reference Figure 1 , a data processing method may include:
[0058] S11. Obtain an initial road book generated based on parameter configuration information and a pre-configured three-dimensional map.
[0059] If the data processing method of the present invention is applied to the off-road field, the parameter configuration information is the off-road parameter configuration information, the three-dimensional map is the three-dimensional off-road map, and the initial road book is the initial three-dimensional road book.
[0060] In practical applications, road books are generally generated manually based on experience. However, due to the limitation of experience, the reliability of manually generated road books is poor. Therefore, in this embodiment, a model can be used. The preferred model can be LLM (Large Language Model) for road book generation. This is because it has powerful language understanding and generation capabilities, and can provide a new solution for the generation of intelligent road books.
[0061] LLM suitable for off-road applications may be, for example, the GPT (Generative Pre-Trained) series, specifically GPT-3, GPT-4, etc., or the BERT (Bidirectional Encoder Representations from Transformers, a pre-trained language model based on the Transformer architecture) series, such as RoBERTa (Robustly Optimized BERT Pretraining Approach, a pre-trained language model based on BERT), ERNIE (Enhanced Representation through Knowledge Integration, a pre-trained language model based on knowledge enhancement), etc., or other models specifically optimized for off-road scenarios. In order to improve the model's adaptability to the off-road field and better output a road book that matches the off-road field, the present invention can also use some off-road training data to train the model to increase the model's knowledge in the off-road field.
[0062] When using the model to generate a road book, it is necessary to refer to the map information of the local off-road scene. Therefore, in this embodiment, a three-dimensional map of the off-road area will be generated in advance. In addition, different users have different requirements for off-road road books. For example, users can configure the starting point and end point according to their own needs to generate a road book. In addition, users can also configure preference information, such as the required difficulty, time consumption, scenery and other personalized needs. When the model generates a road book, it not only considers the starting point and end point set by the user, but also considers the user's preference information, so that the generated initial road book meets the user's needs and the user satisfaction is high.
[0063] S12. In the process of performing activities according to the initial road book, based on the detected front obstacle information, dynamically adjust the initial road book to obtain a first target road book, so as to perform activities based on the first target road book.
[0064] When the user conducts off-road activities, he will first play and climb according to the itinerary in the initial road book. However, in the process of moving forward, he may encounter some obstacle information, such as bad weather such as rain, snow, storm, mudslide, or road collapse, which makes it impossible to continue to move forward according to the initial road book. Therefore, the present invention supports intelligent adaptive adjustment of the road book according to the actual situation ahead during the off-road activity to obtain the first target road book, and the user continues the off-road activity according to the first target road book.
[0065] S13: If a rescue request is received, a target rescue route is determined based on the current rescue data of the activity, so as to perform a rescue operation based on the target rescue route.
[0066] In this embodiment, if the user believes that the obstacle ahead is serious and cannot be avoided even if the route is changed, or if the obstacle ahead causes problems such as getting lost, getting injured, or vehicle failure, or if the user encounters emergencies such as getting lost, getting injured, or vehicle failure during the actual journey, the user can trigger the drone to output a rescue request. After receiving the rescue request, the above-mentioned data processing device will receive the current rescue data collected by the drone, such as the current environment, personnel injuries, food reserves, weather conditions, etc., and then plan a target rescue route and output the target rescue route to the corresponding rescue personnel. The rescue personnel analyze whether the rescue route is feasible. If not, the target rescue route can be manually adjusted, and then the corresponding rescue operation can be performed. If it is feasible, the rescue operation is performed manually according to the target rescue route, so that the people in distress can be rescued as soon as possible to avoid artificial casualties.
[0067] In the present invention, when a user is in distress or needs help, the drone can be used to output a rescue request in a timely manner, so as to plan a rescue route as quickly as possible, perform corresponding rescue operations, and rescue corresponding personnel in a timely manner. Compared with manual self-rescue methods, the rescue efficiency, accuracy and reliability can be improved.
[0068] It should be noted that, during the journey, the route adjustment in step S12 may be omitted and the rescue operation in step S13 may be directly performed.
[0069] In this embodiment, if obstacle information ahead is detected, the initial road book is dynamically adjusted to obtain a first target road book, that is, the present invention can adjust the road book in real time based on the obstacle information, so that the adjusted road book matches the current obstacle situation, which is convenient for users to carry out activities according to the adjusted road book, and the road book adjustment has a high degree of intelligence. In addition, in the present invention, when a rescue request is received, the target rescue route can be determined based on the current rescue data of the activity, so that rescue operations can be performed based on the target rescue route, and rescue services can be provided in a timely manner to avoid casualties. As can be seen from the above, the present invention provides handling measures when encountering obstacles and when rescue is needed, which can improve the degree of intelligence in handling emergencies.
[0070] In another implementation of the present invention, a three-dimensional map needs to be generated in advance so that the environmental data of the area, such as where the road is, where the wetland is, etc., can be provided during the subsequent road book planning. Figure 2 ,The 3D map generation process includes:
[0071] S21. Collect environmental data of the target area.
[0072] In practical applications, it is difficult to manually explore environmental data due to the complexity and changeability of off-road areas. Therefore, in this embodiment, drones are used to collect environmental data. In this embodiment, high-performance drones are selected, equipped with sensor equipment such as LiDAR (Light Detection And Ranging), high-definition cameras, infrared sensors, etc., for autonomous cruising and accurate data collection in off-road environments. The drone should have long endurance, high stability and wind resistance to adapt to the complex and changeable off-road environment.
[0073] During the actual flight process, the drone can conduct all-round, multi-angle scanning and photography of the target area according to the preset cruise path or real-time instructions to obtain detailed information such as terrain, vegetation, obstacles, etc.
[0074] UAVs transmit various sensor data (such as images and point cloud data collected by high-definition cameras, point cloud data collected by lidar, infrared information collected by infrared sensors, etc.) in real time to data processing equipment, such as a data processing center, through communication components. Data collection is accurate and timely.
[0075] In actual applications, the drone can adaptively adjust its altitude according to the local terrain during flight. For example, in a ravine, the altitude can be lowered, and if on a mountain peak, the altitude can be raised to ensure the safety of the drone's operation and prevent obstacles such as trees and peaks from damaging the drone.
[0076] In order to collect comprehensive environmental data in off-road areas, drones can make multiple round trips in the area to collect data comprehensively and avoid the problem of incomplete mapping caused by omissions in data collection.
[0077] S22: Generate an initial three-dimensional map using the environmental data.
[0078] After using drones to collect images, point cloud data and infrared information, since high-definition cameras, lidar and infrared sensors collect different data at the same location, it is necessary to perform data fusion operations on these data to avoid the problem of erroneous collection by a single sensor.
[0079] Before data fusion, the collected data may be preprocessed, such as data cleaning, and then data fusion may be performed to obtain environmental information at different locations in the off-road area.
[0080] Then, a high-precision map generation algorithm is used to generate a map, such as based on SLAM (Simultaneous Localization and Mapping) technology, combined with point cloud data, image information and infrared information collected by drones, to build a high-precision and high-definition initial three-dimensional map. The constructed map not only contains traditional geographic location information, but also includes detailed information such as terrain height, terrain undulation, slope, curvature, vegetation density, vegetation coverage, obstacle distribution, etc., for subsequent path planning.
[0081] In practical applications, infrared information is mainly used to detect living animals such as animals and humans. However, in practical applications, the position of animals can be changed due to their walking. Therefore, when generating the initial three-dimensional map, only point cloud data and image data can be used for map generation operations.
[0082] S23: Mark the type of each target in the initial three-dimensional map to obtain a three-dimensional map.
[0083] In this embodiment, although an initial three-dimensional map is generated, it is only an environmental summary and the positions of key elements such as roads, obstacles, waters, vegetation, etc. in the map cannot be understood. When a cross-country road book is subsequently generated, the route cannot be correctly planned. Therefore, in this embodiment, it is necessary to mark various targets (such as roads, obstacles, waters, vegetation and other key elements) in the initial three-dimensional map to obtain a final three-dimensional map, so that a feasible path can be planned more quickly during subsequent path planning, and when manually viewing the map, the general distribution of key elements in the map can be understood, so that the exit can be found as soon as possible during rescue, and an area that is more suitable for walking can be found when moving forward.
[0084] In another implementation of the present invention, referring to Figure 3 , step S23 may include:
[0085] S31. Perform feature extraction operation on the environmental data to obtain target features.
[0086] In practical applications, in order to more accurately identify the target type, you can first perform feature extraction on the environmental data to extract the target features that can characterize the target type. In order to improve the accuracy of feature extraction, you can use a deep learning model, such as Transformer, BERT, etc.
[0087] S32: Using the target features, perform semantic segmentation on the environmental data to identify various targets in the environmental data.
[0088] In this embodiment, after the target feature is extracted, the target feature can be used to perform a semantic segmentation operation on the environmental data.
[0089] Semantic segmentation refers to assigning each pixel in an image to a corresponding category, which is a classification operation, such as identifying a certain area in environmental data as a road and another area as a water area, etc. In this embodiment, the road, water area, etc. are the targets in this embodiment.
[0090] S33: Mark each target in the initial three-dimensional map to obtain a three-dimensional map.
[0091] In practical applications, when marking targets, the positions of different targets in the initial three-dimensional map are determined, and then different colors or marking types are used to mark them in the initial three-dimensional map. In addition, text prompts of the target types can be specifically marked to enable humans to clearly understand the various components of the off-road area.
[0092] It should be noted that after collecting the environmental data, the steps of generating an initial three-dimensional map and identifying the target's actions can be performed simultaneously, so that after the initial three-dimensional map is generated, the target annotation operation can be performed to obtain a three-dimensional map, thereby improving the efficiency of three-dimensional map generation.
[0093] In addition to using SLAM to generate maps, big data analysis models can also be introduced to conduct in-depth mining and analysis of massive historical off-road data, terrain data, and meteorological data to obtain a comprehensive understanding of off-road terrain characteristics and ultimately construct a three-dimensional map. This high-precision map can provide off-roaders with intuitive and detailed navigation references.
[0094] After obtaining the three-dimensional map, if the user has a need to generate an off-road road book, the three-dimensional map can be used to generate the off-road road book. Figure 4 , obtain the initial road book generated based on parameter configuration information and pre-configured 3D map, including:
[0095] S41. Obtain parameter configuration information.
[0096] Specifically, the user can input parameter configuration information in voice or text on the human-computer interaction interface. The parameter configuration information includes: starting point, end point and preference information. The preference information includes difficulty, time consumption, scenery, etc.
[0097] Difficulty refers to the off-road difficulty selected by the user, such as high difficulty, medium difficulty, low difficulty, etc. Duration refers to the duration of the off-road activity selected by the user, such as one hour, two hours, etc. Scenery refers to the scenic spots that the user wants to see during the off-road activity, such as rivers, mountains, etc. In addition, users can also enter other information according to their needs, such as making the road as flat as possible and providing routes along the river as much as possible.
[0098] S42: Generate a reference road book that meets the parameter configuration information using a large model based on the three-dimensional map.
[0099] The large model selected in this embodiment is the above-mentioned LLM model.
[0100] After receiving the parameter configuration information input by the user, word embedding is used to convert it into a vector, which is used as the input of the LLM model. The LLM model can then generate a reference road book based on user needs.
[0101] The generated road book can be pushed to the human-computer interaction interface. After seeing the road book, the user can output the road book adjustment information by voice or text. For example, if the user thinks the recommended itinerary is too tight, the user can output "Please reduce the itinerary" by voice. After receiving the instruction, the large model will continue to adjust the road book and reduce the itinerary. If the user thinks the arranged route is too difficult, such as needing to climb multiple mountains, the user can output "Please reduce the difficulty" by voice. After receiving the instruction, the model will continue to adjust the road book content, and so on and so forth until a reference road book satisfactory to the user is generated.
[0102] In another implementation of the present invention, in order to facilitate selection, multiple reference road books may be generated for subsequent selection of the optimal road book.
[0103] S43, selecting the best road book from the reference road book to obtain an initial road book.
[0104] In practical applications, path planning algorithms such as the A* algorithm and the Dijkstra algorithm can be used to optimize and sort the reference road books and recommend the best off-road route to the user. For example, the A* algorithm is used to select the best road book with low difficulty or shortest travel time, and the best road book is the initial road book in this embodiment. In addition, the algorithm can also be used to combine historical data and real-time weather information to evaluate the feasibility and safety of the off-road route and select the best road book.
[0105] It should be noted that step S43 can also be implemented through a large model, that is, the large model automatically generates multiple reference road books based on the high-precision three-dimensional map and the starting point, end point and preference settings (such as difficulty, time consumption, scenery, etc.) input by the user. The large model is used to further optimize and sort the reference road books and recommend the best road book to the user.
[0106] In addition, after generating multiple reference road books, all the reference road books can be sent to the user through human-computer interaction, allowing the user to select the best road book that suits him / her best.
[0107] In addition, after generating multiple reference road books, the matching degree between each reference road book and the user may be calculated based on user attribute information (such as user gender, user height, user age, user hobbies, etc.), and the reference road book with the highest matching degree may be recommended to the user.
[0108] In this embodiment, multiple reference road books can be generated, from which the best road book can be selected and provided to the user to meet the user's needs.
[0109] In addition, in the present invention, when generating the initial road book, a large model is used. When generating the road book, the large model will take into account the actual situation of the three-dimensional off-road environment and information such as user needs. Compared with the method of manually relying on experience, paper maps or limited satellite images to generate road books, it can comprehensively and accurately consider complex and changeable terrain, real-time road conditions and user needs, thereby improving the accuracy and feasibility of road book generation.
[0110] In another implementation of the present invention, a specific implementation of "dynamically adjusting the initial road book based on the detected front obstacle information to obtain a first target road book" is provided, which may specifically include the following steps:
[0111] 1) Obtain the obstacle information ahead detected by the drone.
[0112] While the user is traveling according to the initial road book, the drone can perform data collection operations in front of the user to collect real-time off-road conditions and facilitate timely route adjustments.
[0113] The drone can detect forward obstacle information, which may include obstacle information and / or severe weather information.
[0114] Obstacle information may include: a huge rockfall blocking the road ahead, a mudslide, a landslide, etc. In this case, the user can no longer continue to move forward according to the initial road book and needs to change the itinerary.
[0115] In addition, if there is bad weather ahead, such as a tornado or heavy rain, the user cannot continue according to the initial route book and needs to change the itinerary.
[0116] In this embodiment, in addition to being collected through drones, severe weather can also be obtained by obtaining weather forecast information through the Internet. However, weather forecast information may be inaccurate and untimely. Therefore, it is also preferred to use drones to explore the environment ahead.
[0117] Generally speaking, as long as at least one of obstacle information and severe weather information appears, the road book adjustment operation can be performed.
[0118] 2) Using the large model and based on the forward obstacle information, adjusting the content of the initial road book to obtain a first target road book.
[0119] After detecting and obtaining the obstacle information ahead, the large model can be called to re-plan the route based on the obstacle information ahead, the current location, the end point and the preference information to adjust the content of the initial road book and obtain the first target road book.
[0120] When adjusting the road book content, if there is a mudslide ahead, you can bypass it from the nearby peaks, bypass the mudslide area, and then return to the original route to continue driving. Alternatively, you can directly plan a route to the destination from the nearby peaks.
[0121] Unlike conventional path planning, in off-road scenarios, the environment is harsh and you may often have to climb mountains and wade through water, which consumes a lot of manpower and material resources. In conventional path planning, the roads are generally flat and the physical exertion is small. Therefore, when using large models for path planning in off-road environments, it is necessary to fully consider the user's needs for off-road. For example, some users only experience off-road activities with low route difficulty, while some users have rich off-road experience and want to challenge themselves, so they will choose high-difficulty off-road activities.
[0122] In addition to considering user preferences, it is also necessary to consider the physical exertion of off-road routes. For example, experienced users choose difficult off-road activities, but in the actual process, the mountain they have to climb is too difficult, there are many thorns, and there are many fallen rocks, and it is not easy to find a climbing position. They may become exhausted halfway and cannot continue to move forward. It is also difficult to go back and return to the starting point. At this time, they will be in a dilemma.
[0123] In order to avoid this problem, the large model in the present invention not only considers the user's preferences, but also the degree of physical exertion of the trip. When determining the degree of physical exertion of the trip, it is necessary to consider the attribute information of each target that needs to be passed during the trip, and the target can be a mountain, a river, a grassland, etc. The attribute information of the mountain can be such as the height, slope, number of falling rocks, and whether there is a climbing point. The attribute information of the river can be such as the depth and length of the wading. The attribute information of the grassland can be such as: the height of the grass, the type of grass, etc.
[0124] After obtaining the attribute information of the target, it can be input into the model together with the user preference information, the obstacle information ahead, the current location and the end point information, and the optimization operation can be performed in the model to find the appropriate first target road book.
[0125] In addition, in another implementation of the present invention, when determining the first target road book, the location information of the rescue station can also be considered. Generally, the off-road environment is harsh, and physical exhaustion, fainting, and food shortage may occur at any time. If in some convenient off-road, some places will provide rescue stations, which can provide water, food, Band-Aids and other off-road necessities. Therefore, when planning the route, the location information of the rescue station can be considered. If there is a rescue station near the current location of the user, the user can go to the rescue station first and then continue the off-road activity. That is, in the first target road book, the route to the rescue station will be planned first, and then the route to the destination will be planned.
[0126] In addition, in another implementation of the present invention, since the user has been engaged in off-road activities for a period of time and has consumed some physical strength, when planning the first target route book, the difficulty of the trip can be reduced so that the user can reach the destination as soon as possible.
[0127] After the big model plans a new first-destination route book, it can still be pushed to the user for content adjustment operations, so as to generate a route book that satisfies the user through continuous interaction with the user.
[0128] In this embodiment, the off-road environment ahead is collected by drones, and the large model updates the road book content based on the environment ahead in real time, thereby realizing a deep integration of the large model and drone-based technology. In the process of comprehensive and accurate perception and intelligent analysis of the terrain, emergencies that may be encountered during the off-road process (such as mud and rock flow, landslide, etc.) are taken into consideration, and a route adjustment strategy is added to realize real-time acquisition of off-road terrain data and intelligent optimization and adjustment of the off-road route.
[0129] In addition, real-time navigation services are provided, and the routes are dynamically adjusted and optimized based on the road condition information transmitted back by drones in real time to ensure the safety and smoothness of the off-road process.
[0130] Based on this embodiment, in addition to adjusting the road book according to the obstacle information ahead, the road book can also be adjusted based on the user's adjustment demand information.
[0131] For example, when a user follows the initial route book and finds that the off-road route is too simple and not challenging, the user's adjustment requirements can be output by voice or text, such as "Please increase the difficulty of the route". Alternatively, if the user finds that the route is too difficult and wastes physical energy, the user can output by voice "Please reduce the difficulty of the route". Alternatively, if the user finds that rock climbing alone is too monotonous, the user can output by voice "Please recommend a route with a river".
[0132] After obtaining the user adjustment requirement information, the initial road book may be dynamically adjusted based on the user adjustment requirement information to obtain a second target road book, so as to perform off-road activities based on the second target road book.
[0133] In specific implementation, similar to the generation of the first target road book mentioned above, a large model can also be used to adjust the road book content to generate a second target road book that meets user needs and recommend it to the user so that the user can continue to move forward according to the content.
[0134] As the user moves forward, the road book content can be continuously adjusted according to user needs or obstacle information until the end of the journey.
[0135] In the process of performing off-road activities according to the initial road book, in order to prompt the user's current position and direction of travel, the location information of the current off-road activity can be detected by using GPS (Global Positioning System) and other methods, and then navigation information is determined based on the location information.
[0136] Specifically, based on the current position information, the position information can be determined from the three-dimensional map, and then the next driving location of the current position information in the initial road book is checked, and the navigation information is determined according to the current position information and the relative position of the next driving location. Then, the voice output includes the position information and the semantic prompt information of the navigation information, so that the user can understand the current position and travel direction in a timely, clear and accurate manner.
[0137] Among them, in order to realize the intelligent output of navigation information, the interactive method of AR (Augmented Reality) navigation can also be adopted, so that users can view the direction of travel more directly.
[0138] In this embodiment, during the off-road process, combined with the road condition information (such as sudden obstacles, weather changes, etc.) transmitted back by the drone in real time, the route is dynamically adjusted and optimized according to user needs and / or real-time road conditions, thereby achieving adaptive adjustment of the route to avoid risks and improve personnel safety.
[0139] In another implementation of the present invention, if a user needs rescue, a rescue request can be output through a drone, and then the data processing device will plan a rescue route and perform rescue operations in a timely manner. In this embodiment, if a communication device such as a mobile phone has a signal, a rescue request can also be output by making a phone call, sending a text message, etc.
[0140] Then, refer to Figure 5 , "Determine the target rescue route based on the current rescue data of the activity", including:
[0141] S51. Obtain the current rescue data collected by the drone.
[0142] In this embodiment, if the user believes that the obstacle ahead is serious and cannot be avoided even by changing the route, or if the obstacle ahead causes getting lost, injury, vehicle breakdown, and other problems, or if the user encounters getting lost, injury, vehicle breakdown, and other emergencies during the actual journey, the user can trigger the drone to output a rescue request. After receiving the rescue request, the above-mentioned data processing device will receive the current rescue data collected by the drone, such as the current environment, personal injury status, food reserves, weather conditions, and other data, and use the collected data for subsequent rescue operations.
[0143] In addition, drones can be used for aerial reconnaissance and search and rescue, while sending location information and on-site video to rescue teams.
[0144] S52: planning a reference rescue route based on the current rescue data.
[0145] In practical applications, the big model can be used to analyze the current rescue data and plan a reference rescue route. For example, when planning a route, the nearest exit can be analyzed, and then a reference rescue route from the current location to the exit can be planned.
[0146] S53: Analyze the rescue risk of the reference rescue route.
[0147] Specifically, rescue risk can be realized by using large models and machine learning algorithms. Specifically, it can analyze and evaluate multiple factors such as the feasibility, safety, time consumption, driving time, driving difficulty of the reference rescue route, and the relative position of the route to the rescue personnel to obtain rescue risk and provide a scientific basis for route planning. Rescue risk can be divided into high, medium and low types.
[0148] S54: Select a target rescue route from the reference rescue routes, the rescue risk of which satisfies a preset rescue rule.
[0149] Specifically, the preset rescue rule may be such as the lowest risk, which can ensure the safety of personnel as much as possible. Therefore, after obtaining the rescue risk, the route with the lowest rescue risk is selected as the target rescue route.
[0150] In addition, when determining the target rescue route, in addition to considering the rescue risk, the rescue difficulty can also be considered. For example, if a user falls into a ravine, the rescue is difficult, more rescue equipment will be required, and the rescue efficiency requirements are high. At this time, a drone may be needed to airdrop for rescue. If the user is just lost but there are no casualties, the rescue difficulty is low, and only simple equipment is needed to achieve rescue, and rescue can be carried out by car or on foot.
[0151] After obtaining the target rescue route, the rescuers can adjust the route based on their experience. For example, if the route requires climbing a mountain, but the person's leg is injured and it is not suitable for climbing the mountain, then a detour can be taken. After determining the rescue route, the rescuers are organized to carry out the rescue operation.
[0152] During the rescue process, drones can be used to monitor the safety of personnel and transmit video in real time to view the rescue progress in real time.
[0153] In this embodiment, the rescue difficulty and risk are evaluated to analyze the best rescue path to ensure the efficiency and safety of the rescue operation.
[0154] In addition, the embodiments of the present invention use a large model to plan a reference rescue route based on the data collected by the drone in real time, and provide accurate decision-making support for the rescue team. When encountering emergencies (such as getting lost, injured, vehicle failure, etc.) during the off-road process, the system can respond quickly and initiate emergency plans and auxiliary decisions, thereby improving the safety and rescue efficiency of off-road activities.
[0155] As can be seen from the above, the present invention provides a strategy for generating off-road route books based on large models and drone assistance. By integrating advanced large models, drone autonomous cruising and data collection technology, and high-precision map generation algorithms, it provides accurate and real-time route planning and navigation services for off-road enthusiasts, expedition teams, and emergency response teams.
[0156] Based on the above data processing method embodiment, another embodiment of the present invention provides a data processing device, referring to Figure 6 , which may include:
[0157] A road book acquisition module 11 is used to acquire an initial road book generated based on parameter configuration information and a pre-configured three-dimensional map;
[0158] A road book adjustment module 12 is used to dynamically adjust the initial road book based on the detected front obstacle information during the process of performing activities according to the initial road book to obtain a first target road book, so as to perform activities based on the first target road book;
[0159] The rescue control module 13 is used to determine a target rescue route based on the current rescue data of the activity if a rescue request is received, so as to perform a rescue operation based on the target rescue route.
[0160] In one implementation, a map generation module is also included, and the map generation module includes:
[0161] A data collection submodule is used to collect environmental data of the target area;
[0162] A map generation submodule, used to generate an initial three-dimensional map using the environmental data;
[0163] The map marking submodule is used to mark each target in the initial three-dimensional map to obtain a three-dimensional map.
[0164] In one implementation, the map annotation submodule includes:
[0165] A feature extraction unit, used to perform a feature extraction operation on the environmental data to obtain target features;
[0166] A semantic segmentation unit, configured to perform a semantic segmentation operation on the environmental data using the target features to identify various targets in the environmental data;
[0167] The marking unit is used to mark each target in the initial three-dimensional map to obtain a three-dimensional map.
[0168] In one implementation, the road book acquisition module 11 includes:
[0169] An information acquisition submodule is used to acquire parameter configuration information, wherein the parameter configuration information includes: a starting point, an end point, and preference information;
[0170] A road book generation submodule, used to generate a reference road book that conforms to the parameter configuration information based on the three-dimensional map using a large model;
[0171] The road book screening submodule is used to screen out the best road book from the reference road book to obtain an initial road book.
[0172] In one implementation, the road book adjustment module 12 is specifically used for:
[0173] Obtaining the front obstacle information detected by the drone; the front obstacle information includes obstacle information and / or severe weather information; using the large model to adjust the content of the initial road book based on the front obstacle information to obtain a first target road book.
[0174] In one implementation, the rescue control module 13 includes:
[0175] The data acquisition submodule is used to obtain the current rescue data collected by the drone;
[0176] A rescue route planning submodule, used for planning a reference rescue route based on the current rescue data;
[0177] A risk analysis submodule, used to analyze the rescue risk of the reference rescue route;
[0178] The route screening submodule is used to select a target rescue route whose rescue risk meets the preset rescue rules from the reference rescue routes.
[0179] In one implementation, the road book adjustment module 12 is further configured to:
[0180] Acquire user adjustment demand information; based on the user adjustment demand information, dynamically adjust the initial road book to obtain a second target road book, and perform activities based on the second target road book.
[0181] In one implementation, the method further includes:
[0182] a data determination module, configured to detect location information of a current activity and determine navigation information based on the location information;
[0183] The prompt information output module is used to output voice prompt information, and the voice prompt information includes the location information and the navigation information.
[0184] In this embodiment, if obstacle information ahead is detected, the initial road book is dynamically adjusted to obtain a first target road book, that is, the present invention can adjust the road book in real time based on the obstacle information, so that the adjusted road book matches the current obstacle situation, which is convenient for users to carry out activities according to the adjusted road book, and the road book adjustment has a high degree of intelligence. In addition, in the present invention, when a rescue request is received, the target rescue route can be determined based on the current rescue data of the activity, so that rescue operations can be performed based on the target rescue route, and rescue services can be provided in a timely manner to avoid casualties. As can be seen from the above, the present invention provides handling measures when encountering obstacles and when rescue is needed, which can improve the degree of intelligence in handling emergencies.
[0185] It should be noted that, for the working process of each module, sub-module and unit in this embodiment, please refer to the corresponding description in the above embodiment, which will not be repeated here.
[0186] The present application also provides an electronic device in an embodiment. The electronic device is the above-mentioned off-road data processing device. Figure 7 As shown, it shows a schematic diagram of the structure of an electronic device suitable for implementing the embodiments of the present application. The electronic device in the embodiments of the present application may include but is not limited to fixed terminals such as mobile phones, laptops, PDAs (personal digital assistants), PADs (tablet computers), desktop computers, etc., and may also be a data processing center composed of a single or multiple devices. Figure 7 The electronic device shown is merely an example and should not bring any limitation to the functions and scope of use of the embodiments of the present application.
[0187] like Figure 7As shown, the electronic device may include a processing device (e.g., a central processing unit, a graphics processor, etc.) 601, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 602 or a program loaded from a storage device 608 to a random access memory (RAM) 603 to implement an off-road data processing method. When the electronic device is powered on, various programs and data required for the operation of the electronic device are also stored in the RAM 603. The processing device 601, the ROM 602, and the RAM 603 are connected to each other via a bus 604. An input / output (I / O) interface 605 is also connected to the bus 604.
[0188] Typically, the following devices may be connected to the I / O interface 605: an input device 606 including, for example, a touch screen, a touch pad, a keyboard, a mouse, a camera, a microphone, an accelerometer, a gyroscope, etc.; an output device 607 including, for example, a liquid crystal display (LCD), a speaker, a vibrator, etc.; a storage device 608 including, for example, a memory card, a hard disk, etc.; and a communication device 609. The communication device 609 may allow the electronic device to communicate with other devices wirelessly or by wire to exchange data. Although Figure 7 An electronic device having various devices is shown, but it should be understood that it is not required to implement or possess all the devices shown. More or fewer devices may be implemented or possessed instead.
[0189] An embodiment of the present application also provides a computer program product including computer-readable instructions. When the computer-readable instructions are executed on an electronic device, the electronic device implements any off-road data processing method provided in the embodiment of the present application.
[0190] A computer-readable storage medium is also provided in an embodiment of the present application. The storage medium carries one or more computer programs. When the one or more computer programs are executed by an electronic device, the electronic device can implement any off-road data processing method provided in the embodiment of the present application.
[0191] The above description of the disclosed embodiments enables one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.
Claims
1. A data processing method, characterized in that: include: Obtain an initial road book generated based on parameter configuration information and a pre-configured three-dimensional map; In the process of performing activities according to the initial road book, based on the detected front obstacle information, the initial road book is dynamically adjusted to obtain a first target road book, so as to perform activities based on the first target road book; If a rescue request is received, a target rescue route is determined based on the active current rescue data, so as to perform a rescue operation based on the target rescue route.
2. The data processing method according to claim 1, characterized in that: The generation process of the three-dimensional map includes: Collect environmental data of the target area; Using the environmental data, generating an initial three-dimensional map; Each target is marked in the initial three-dimensional map to obtain a three-dimensional map.
3. The data processing method according to claim 2, characterized in that: Marking each target in the initial three-dimensional map to obtain a three-dimensional map includes: Performing a feature extraction operation on the environmental data to obtain target features; Using the target features, performing a semantic segmentation operation on the environmental data to identify various targets in the environmental data; The various targets are marked in the initial three-dimensional map to obtain a three-dimensional map.
4. The data processing method according to claim 1, characterized in that: Get the initial road book generated based on parameter configuration information and pre-configured 3D map, including: Obtain parameter configuration information, the parameter configuration information including: a starting point, an end point, and preference information; Based on the three-dimensional map, a reference road book conforming to the parameter configuration information is generated using a large model; The best road book is selected from the reference road book to obtain an initial road book.
5. The data processing method according to claim 1, characterized in that: Based on the detected obstacle information ahead, dynamically adjusting the initial road book to obtain a first target road book includes: Obtaining front obstacle information detected by the drone; the front obstacle information includes obstacle information and / or severe weather information; The large model is used to adjust the content of the initial road book based on the front obstacle information to obtain a first target road book.
6. The data processing method according to claim 1, characterized in that: Determine the target rescue route based on the current rescue data of the activity, including: Obtain the current rescue data collected by the drone; Based on the current rescue data, planning a reference rescue route; Analyzing the rescue risk of the reference rescue route; A target rescue route whose rescue risk meets a preset rescue rule is selected from the reference rescue routes.
7. The data processing method according to claim 1, characterized in that: In the process of carrying out activities according to the initial road book, it also includes: Obtain user adjustment demand information; Based on the user adjustment demand information, the initial road book is dynamically adjusted to obtain a second target road book, so as to carry out activities based on the second target road book.
8. The data processing method according to claim 1, characterized in that: In the process of carrying out activities according to the initial road book, it also includes: Detect the location information of the current activity; determining navigation information based on the location information; Output voice prompt information, where the voice prompt information includes the location information and the navigation information.
9. A data processing device, characterized in that: include: A road book acquisition module is used to acquire an initial road book generated based on parameter configuration information and a pre-configured three-dimensional map; A road book adjustment module is used to dynamically adjust the initial road book based on the detected front obstacle information during the process of performing activities according to the initial road book to obtain a first target road book, so as to perform activities based on the first target road book; The rescue control module is used to determine a target rescue route based on the current rescue data of the activity if a rescue request is received, so as to perform a rescue operation based on the target rescue route.
10. An electronic device, characterized in that: The method comprises at least one processor and a memory connected to the processor, wherein: The memory is used to store computer programs; The processor is used to execute the computer program so that the electronic device can implement the data processing method according to any one of claims 1 to 8.
