Method for spatial arrangement of a vehicle trunk and vehicle

Abstract

This application discloses a method for spatial layout of a vehicle trunk and a vehicle in the field of intelligent cockpit technology. When the currently available space in the trunk is insufficient to accommodate items to be loaded, the method combines the actual occupancy status of the rear seats with the optimization goal of maximizing rear passenger space while meeting the needs for item storage. It then formulates a reasonable rear seat adjustment scheme and determines the corresponding target available space, achieving dynamic adaptive expansion of the storage space. Based on this, and combining the target available space and the spatial attribute parameters of the items to be loaded, an adaptive loading layout strategy is further generated, and the seat adjustment scheme and loading layout strategy are output uniformly to guide the user through the operation. This method deeply integrates dynamic space adjustment with loading planning under multiple constraints, fully exploring the potential of in-vehicle storage space while balancing loading and passenger usage needs.

Description

Technical Field

[0001] This application relates to the field of intelligent cockpit technology, and in particular to a spatial layout method for a vehicle trunk and a vehicle thereof. Background Technology

[0002] With the continuous evolution of automotive intelligent technology, users are increasingly demanding higher levels of convenience, space utilization, and intelligent interaction in vehicle storage scenarios. As the main loading space of a vehicle, the user experience of the trunk directly affects the overall driving experience and product market competitiveness. Intelligent and adaptive in-vehicle storage spaces have become the mainstream development direction of storage technology in the automotive industry.

[0003] Currently, the industry has developed technologies related to matching storage space with cargo and autonomous space adjustment. For example, an automatic trunk volume adjustment system for automobiles can detect the remaining volume in each storage area, select suitable areas for loading items, and alert the user. When the existing space cannot accommodate the items, it can drive the partition to move, dynamically changing the storage space volume to complete the loading operation. Furthermore, the aviation industry has mature loading planning technologies that can optimize cargo loading schemes within a fixed cabin space by combining constraints such as weight balance and space matching. Based on these two types of technologies, a technical approach can be proposed that first senses the volume of objects, dynamically adjusts the vehicle's storage space via partitions, and then conducts loading planning. However, this technical concept fails to fully explore the potential of vehicle storage space and cannot simultaneously meet the comprehensive needs of dynamic adaptive adjustment of storage space and intelligent loading planning under multiple constraints, resulting in significant limitations in overall functionality and practicality. Summary of the Invention

[0004] In view of the above problems, this application provides a method for spatial layout of a vehicle trunk and a vehicle that overcomes or at least partially solves the above problems. The technical solution is as follows: A method for spatial layout of a vehicle trunk, comprising: when the current available space in the vehicle trunk cannot accommodate the items to be loaded, generating a rear seat adjustment scheme and a corresponding target available space based on the occupancy status of the rear seats, with the optimization objective of satisfying the accommodating needs of the items to be loaded and maximizing the remaining rear seating space; generating a loading layout strategy for the items to be loaded based on the target available space and the spatial attribute parameters of the items to be loaded; and outputting the rear seat adjustment scheme and the loading layout strategy to guide the user to adjust the rear seats and place the items to be loaded in the trunk.

[0005] This application addresses situations where the available space in the trunk is insufficient to accommodate items. It considers the actual occupancy of the rear seats to optimize both cargo capacity and rear passenger space. A reasonable rear seat adjustment plan is developed, and the corresponding target available space is determined, enabling dynamic adaptive expansion of storage space. Based on this, and combining the target available space with the spatial attribute parameters of the items to be loaded, an adaptive loading layout strategy is generated. This strategy provides a unified output of both the seat adjustment plan and the loading layout strategy, guiding the user through the process. This method deeply integrates dynamic space adjustment with loading planning under multiple constraints, fully exploring the potential of in-vehicle storage space while balancing loading and passenger comfort.

[0006] Optionally, before generating a rear seat adjustment scheme and corresponding target available space based on the occupancy status of the rear seats of the vehicle, with the optimization goal of satisfying the accommodating needs of the items to be loaded and maximizing the remaining rear seating space, when the current available space in the vehicle's trunk cannot accommodate the items to be loaded, the method further includes: obtaining spatial attribute parameters of the items to be loaded and a first spatial model of the currently available space, wherein the first spatial model is a three-dimensional spatial model of preset components related to the trunk volume in a first state parameter; generating an initial placement strategy for the first spatial model based on the spatial attribute parameters; simulating the placement of the items to be loaded within the first spatial model according to the initial placement strategy, and determining whether the current available space corresponding to the first spatial model can accommodate the items to be loaded.

[0007] The above optional embodiments pre-acquire the spatial attribute parameters of the items to be loaded and the first spatial model corresponding to the currently available space in the trunk. An initial placement strategy is generated based on the item's spatial attributes, and the carrying capacity is determined by simulating item placement in the first spatial model. This method can accurately predict the loading capacity of the existing storage space before performing seat adjustment operations, avoiding meaningless space adjustment actions. Simultaneously, the visualization and precision of loading prediction are achieved through a three-dimensional spatial model and simulated placement, providing reliable data for subsequent seat configuration generation and space expansion, thus improving the overall predictive efficiency and execution rationality of the plan.

[0008] Optionally, obtaining the spatial attribute parameters of the item to be loaded and the first spatial model of the currently available space includes: obtaining image data and depth information of the item to be loaded; the depth information is used to characterize the mapping relationship between two-dimensional pixel coordinates and three-dimensional spatial distance in the image data; identifying the item to be loaded based on the image data and the depth information to obtain the spatial attribute parameters; obtaining the first state parameter of the preset component; determining the first spatial model corresponding to the first state parameter from a preset spatial model library; wherein, the preset spatial model library includes the space model of the trunk corresponding to the preset component under different state parameters.

[0009] The above-mentioned optional embodiments, by collecting image data and depth information of the items to be loaded, rely on these two to complete item identification and extract spatial attribute parameters, accurately reconstructing the three-dimensional size features of the items and ensuring the accuracy of parameter collection. Simultaneously, by reading preset component status parameters, the corresponding 3D model of the storage compartment is retrieved from a pre-set spatial model library, quickly constructing a first spatial model of the currently available space in the trunk. The overall data acquisition and model retrieval process is highly automated and has a fast response speed, achieving both high-precision acquisition of item spatial information and ensuring the efficiency and adaptability of storage space model retrieval, providing realistic and reliable 3D data support for subsequent simulation placement and space capacity determination.

[0010] Optionally, the spatial attribute parameters include: size information, category, contour information, occupied space, and attribute label; the step of identifying the item to be loaded based on the image data and the depth information to obtain the spatial attribute parameters includes: based on the image data and the depth information, calling a three-dimensional vision measurement algorithm to identify the contour edge of the item to be loaded to obtain the size information of the item to be loaded; and performing target detection and instance segmentation on the image data based on a pre-trained convolutional neural network model to obtain the category, contour information, occupied space, and attribute label of the item to be loaded.

[0011] The above optional embodiments combine 3D vision measurement algorithms with pre-trained convolutional neural network models to extract various spatial attribute parameters of the items to be loaded step by step. The 3D vision measurement algorithm accurately identifies the outline edges of the items, reliably obtaining dimensional information; the convolutional neural network is used to complete target detection and instance segmentation, efficiently obtaining item category, outline information, occupied space, and attribute labels. This combined approach can comprehensively and meticulously extract multi-dimensional parameters, with high parameter detection accuracy and excellent recognition efficiency. It can completely restore the spatial and attribute characteristics of the items to be loaded, providing comprehensive and accurate data support for subsequent processes such as space simulation placement, capacity assessment, space adjustment, and loading planning.

[0012] Optionally, the step of simulating the placement of the item to be loaded within the first space model according to the initial placement strategy, and determining whether the currently available space corresponding to the first space model can accommodate the item to be loaded, includes: unifying the physical coordinate system and performing three-dimensional collision detection and simulated placement based on the rule bounding box model of the item to be loaded included in the initial placement strategy and the first space model; and determining whether the currently available space can accommodate the item to be loaded based on the three-dimensional collision detection results and simulation results.

[0013] The above optional embodiments employ a regular bounding box model combined with a unified physical coordinate system to conduct 3D collision detection and simulated placement, relying on 3D spatial calculations to determine the storage space's capacity. This method can realistically reproduce the placement of items within the storage space, accurately detect spatial interference, and effectively improve the accuracy of the capacity judgment. Simultaneously, based on a standardized model and coordinate system, the simulation process is logically rigorous and computationally efficient, quickly outputting judgment results and providing an objective and reliable basis for subsequently initiating the trunk space adjustment process.

[0014] Optionally, the step of unifying the physical coordinate system and performing three-dimensional collision detection and simulated placement based on the rule bounding box model of the item to be loaded included in the initial placement strategy and the first spatial model includes: determining different placement postures for the rule bounding box model and calculating the corresponding three-dimensional dimension parameters under different placement postures; partitioning the first spatial model and determining the placement anchor point positions; sorting the rule bounding box model according to a preset priority, and performing trial placement sequentially according to the placement anchor point positions.

[0015] The above-mentioned optional embodiments, by determining different placement postures of items and calculating corresponding three-dimensional dimensional parameters, can fully explore the various placement possibilities of items in the trunk, avoiding space waste caused by a single posture. At the same time, by dividing the trunk into zones and determining placement anchor points, and by sorting the item models according to preset priorities and trying them out in sequence, the placement of items can be made more orderly and logical, reducing spatial conflicts caused by disordered placement. Furthermore, by eliminating infeasible solutions through three-dimensional collision detection and comprehensively scoring and selecting the best effective placement solution, the initial placement strategy with the highest space utilization and the most reasonable layout can be obtained, thereby improving the trunk space utilization rate, avoiding item interference and space redundancy, providing scientific and efficient layout guidance for subsequent actual loading, and further improving the reliability and practicality of intelligent loading.

[0016] Optionally, the step of simulating the placement of the item to be loaded within the first space model according to the initial placement strategy, and determining whether the currently available space corresponding to the first space model can accommodate the item to be loaded, includes: mapping the spatial attribute parameters of the item to be loaded to the physical coordinate system to which the first space model belongs according to the initial placement strategy, and determining the placement area of ​​the item to be loaded; comparing and verifying the placement area with the boundary of the first space model, and determining whether the placement position of the item to be loaded exceeds the constraint range of the currently available space model; if yes, then determining that the first space cannot accommodate the item to be loaded; if no, then determining that the currently available space can accommodate the item to be loaded.

[0017] The above optional embodiments map the spatial attribute parameters of the items to be loaded onto a unified physical coordinate system of the storage space, accurately delineating the item placement area, and completing the capacity judgment by comparing this area with the boundary of the space model. This judgment logic is simple and clear, with low computational load, and can quickly complete range verification and result judgment. At the same time, relying on the unified coordinate system ensures the accuracy of position comparison, effectively avoiding judgment deviation, and can efficiently and accurately identify whether the current storage space meets the loading requirements, providing a reliable basis for triggering subsequent processes.

[0018] Optionally, after obtaining the spatial attribute parameters of the item to be loaded and the first spatial model of the currently available space, the method further includes: determining the target spatial model required to accommodate the item to be loaded based on the spatial attribute parameters; comparing the size relationship between the target spatial model and the first spatial model in the physical coordinate system to which the first spatial model belongs; and determining that the currently available space cannot accommodate the item to be loaded if the target spatial model is larger than the first spatial model.

[0019] The above optional embodiments construct a corresponding target space model based on the spatial attribute parameters of the items to be loaded. By directly comparing the size relationship between the target space model and the first space model under a unified physical coordinate system, it is possible to quickly determine whether the trunk space meets the loading requirements in an intuitive and quantitative way, simplifying the space adaptability judgment process and effectively improving the efficiency of capacity feasibility detection. By directly comparing the model dimensions, the computational overhead caused by complex simulation placement and posture calculation is avoided, while ensuring the accuracy and reliability of the judgment results and reducing misjudgments caused by estimation errors. Loading schemes with insufficient space can be quickly eliminated, providing a clear basis for subsequent adjustment of item placement posture, optimization of loading strategies, or prompting users to change loading methods, further improving the response speed and judgment accuracy of intelligent planning of the vehicle trunk.

[0020] Optionally, when the current available space in the vehicle's trunk cannot accommodate the items to be loaded, based on the occupancy status of the rear seats, and with the optimization objective of maximizing the remaining rear seating space while meeting the occupancy requirements of the items to be loaded, a rear seat adjustment scheme and a corresponding target available space are generated. This includes: based on the occupancy status of the rear seats, determining a second space model from a preset space model library whose volume is greater than or equal to the target space model when the remaining rear seating space is maximized; the preset space model library includes the trunk space models corresponding to the preset components under different state parameters; generating the rear seat adjustment scheme based on the difference between the second state parameter corresponding to the second space model and the first state parameter, and determining the target available space corresponding to the second space model; the rear seat adjustment scheme is used to guide the user to adjust the first state parameter of the preset component to the second state parameter.

[0021] The above optional embodiment first determines the required target space model based on the spatial attribute parameters of the items to be loaded. Then, combined with the occupancy status of the rear seats, it selects a second space model from a preset space model library that meets the loading requirements and maximizes the remaining rear seating space. This determines the target available space, and a seat adjustment scheme is generated based on the differences in the front and rear state parameters of the components. This method is optimized based on the dual constraints of loading and passenger needs, and relies on a standardized space model library to complete rapid matching. It ensures that the storage space can smoothly accommodate items while maximizing the preservation of rear seating functionality, balancing loading practicality and passenger comfort. At the same time, it generates adjustment schemes based on the differences in state parameters, making it easy for users to perform operations. The space allocation scheme is reasonable and highly adaptable.

[0022] Optionally, generating a loading layout strategy for the items to be loaded based on the target available space and the spatial attribute parameters of the items to be loaded includes: generating a loading layout strategy for the items to be loaded according to the spatial attribute parameters in the physical coordinate system to which the second spatial model of the target available space belongs, wherein the loading layout strategy includes the placement order, placement position and placement posture of the items to be loaded in the second space.

[0023] The above optional embodiments generate a loading layout strategy under the unified physical coordinate system of the second spatial model corresponding to the target available space, combined with the spatial attribute parameters of the items to be loaded, clarifying the placement order, position, and posture of the items. Relying on a unified coordinate system for layout planning ensures a high degree of matching between the layout scheme and the actual storage space, resulting in accurate and reasonable layout planning. Complete layout information can intuitively guide users in placing items, improving the convenience of loading operations and space utilization, and further realizing an effective combination of adaptive adjustment of storage space and intelligent loading planning.

[0024] A space layout device for a vehicle trunk, the device comprising: The adjustment scheme generation module is used to generate a rear seat adjustment scheme and the corresponding target available space based on the occupancy status of the rear seats of the vehicle when the current available space in the trunk of the vehicle cannot accommodate the items to be loaded, with the optimization goal of maximizing the remaining rear seat space while meeting the accommodating requirements of the items to be loaded. The layout strategy generation module generates a loading layout strategy for the item to be loaded based on the target available space and the spatial attribute parameters of the item to be loaded. The output module outputs the rear seat adjustment scheme and the loading layout strategy to guide the user to adjust the rear seat and place the items to be loaded in the trunk.

[0025] A computer-readable storage medium having a program stored thereon, which, when executed by a processor, implements the spatial layout method for a vehicle trunk as described above.

[0026] A vehicle includes a memory and a processor, wherein the memory stores executable program code, and the processor is used to call and execute the executable program code to perform a spatial layout method for the vehicle's trunk.

[0027] A computer program product includes a computer program that, when executed by a processor, implements the steps of any of the above-described methods for the spatial layout of a vehicle trunk.

[0028] The above description is only an overview of the technical solution of this application. In order to better understand the technical means of this application and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of this application more obvious and understandable, the following are specific embodiments of this application. Attached Figure Description

[0029] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit the scope of this application. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings: Figure 1 This is a schematic flowchart illustrating a spatial layout method for a vehicle trunk provided in an embodiment of this application; Figure 2 This is a schematic structural diagram of a space layout device for a vehicle trunk provided in an embodiment of this application; Figure 3 This is a schematic structural diagram of a vehicle provided in an embodiment of this application. Detailed Implementation

[0030] Exemplary embodiments of the present application will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present application are shown in the drawings, it should be understood that the present application may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this application will be thorough and complete, and will fully convey the scope of the present application to those skilled in the art.

[0031] Currently, relevant technologies have been developed to match and adjust the volume of loaded items with storage space. Automatic trunk volume adjustment systems can compare the remaining volume of multiple storage spaces in the trunk, select suitable storage areas for the items to be loaded, and prompt the user. When existing storage space cannot accommodate the items, the system can also move partitions to actively adjust the volume of the storage space to complete the loading. There are also technologies for passenger aircraft baggage compartments that combine weight balance and space matching requirements to plan cargo loading and optimize loading schemes within fixed compartment spaces.

[0032] With the aforementioned technologies already in place, a solution can be derived by first sensing the volume of objects and dynamically adjusting the size of the vehicle's storage space via moving partitions, and then planning how to load them. However, this approach does not fully utilize the potential of the vehicle's storage space and struggles to simultaneously address the combined needs of dynamic space adaptive adjustment and intelligent loading planning under multiple constraints. Furthermore, it lacks a systematic and intelligent loading guidance system from the user's perspective, resulting in low item placement efficiency, significant space waste, and a high likelihood of insufficient space or the tailgate failing to close properly, thus failing to meet users' demands for intelligent and convenient vehicle loading.

[0033] To address some or all of the aforementioned technical problems, this application provides a spatial layout method for a vehicle trunk. This method optimizes and improves upon the shortcomings of related technologies that cannot simultaneously consider dynamic adaptive adjustment of storage space and intelligent loading planning under multiple constraints. When the currently available space in the trunk is insufficient to accommodate the items to be loaded, the method considers the actual occupancy status of the rear seats, aiming to meet the needs of item storage while maximizing the preservation of rear passenger space. A reasonable rear seat adjustment scheme is formulated, and the corresponding target available space is determined, achieving dynamic adaptive expansion of storage space. Based on this, by combining the target available space and the spatial attribute parameters of the items to be loaded, an adaptive loading layout strategy is further generated, and a unified output of the seat adjustment scheme and loading layout strategy guides the user through the operation. This method deeply integrates dynamic space adjustment with loading planning under multiple constraints, fully exploring the potential of in-vehicle storage space while considering both loading and passenger usage needs, effectively solving the problems of limited functionality and insufficient comprehensive adaptability of related technologies.

[0034] like Figure 1As shown, Figure 1 This is a schematic flowchart illustrating a method for spatial layout of a vehicle trunk according to an embodiment of this application. The method includes the following steps S101~S104: S101. If the current available space in the vehicle's trunk cannot accommodate the items to be loaded, based on the occupancy status of the rear seats, with the optimization goal of maximizing the remaining rear seating space while meeting the needs of the items to be loaded, a rear seat adjustment scheme and the corresponding target available space are generated.

[0035] In some embodiments, before executing step S101, the spatial attribute parameters of the items to be loaded in the trunk and the first spatial model of the trunk are first obtained. Then, based on the spatial attribute parameters, an initial placement strategy for the first spatial model is generated; according to the initial placement strategy, the items to be loaded are simulated to be placed in the first spatial model, and it is determined whether the currently available space corresponding to the first spatial model can accommodate the items to be loaded.

[0036] The spatial attribute parameters of the items to be loaded include, but are not limited to: volume, size information, category, outline information, occupied space, and attribute labels. These spatial attribute parameters can be obtained by collecting image data and depth point cloud data of the items to be loaded at the rear of the vehicle, and then jointly analyzing them through 3D vision measurement and instance segmentation. The first spatial model is a real-time, usable 3D spatial model of the trunk in its current state, showing preset components related to the trunk volume (such as the rear seats and trunk panels).

[0037] In the process of obtaining the spatial attribute parameters of the item to be loaded, image data and depth information of the item are first acquired. Then, based on the image data and depth information, the spatial attribute parameters are obtained by identifying the item. The depth information is used to characterize the mapping relationship between two-dimensional pixel coordinates and three-dimensional spatial distances in the image data.

[0038] Optionally, based on image data and depth information, a 3D vision measurement algorithm is invoked to identify the contour edges of the item to be loaded, thereby obtaining the size parameters of the item to be loaded; based on a pre-trained convolutional neural network model, target detection and instance segmentation are performed on the image data to obtain the category, contour information, occupied space, and attribute labels of the item to be loaded.

[0039] The pre-trained convolutional neural network model can use methods such as single-target positioning, binocular stereo vision, or structured light measurement to calculate the three-dimensional dimensions, volume, and space occupied by the item to be loaded, while extracting the category, outline, shape, shape information, and attribute labels of the item to be loaded.

[0040] For example, during a preset time period when the user opens the trunk and holds the items to be loaded, a surround-view camera deployed inside the tailgate or rear bumper continuously collects single or multiple frames of image data containing all the items to be loaded. Simultaneously, it works in conjunction with an onboard depth camera, such as a Time-of-Flight (ToF) camera, or with an ultrasonic sensor to acquire the depth information of the items, establishing a mapping relationship between the two-dimensional pixel coordinates of the items and their three-dimensional spatial distance. Based on the collected image data and depth information, a vision algorithm module is invoked. Depending on the actual hardware configuration, one or more combinations of monocular vision and reference object calibration, binocular stereo vision matching, or structured light depth measurement are selected. Through image processing steps such as feature point extraction, sub-pixel-level edge detection, stereo matching, and depth map reconstruction, the contour edges of each item are identified and located one by one, and then the length, width, and height three-dimensional dimensions of each item are calculated and output.

[0041] The acquired object images are input into a pre-trained convolutional neural network model, which integrates object detection and instance segmentation functions. The model first extracts multi-scale feature maps from the images through a backbone network, then uses the detection head to output the category probability and bounding box information of each target object in the image, enabling the recognition of object categories such as suitcases, cardboard boxes, backpacks, and balls. Subsequently, based on the instance segmentation branch, pixel-level contours of each target object are extracted to generate corresponding mask data, representing its outer contour and occupied space shape. Simultaneously, according to preset category attribute mapping rules, the model associates identified items with preset special attribute labels such as fragile, non-stackable, and priority placement for heavy objects. These mapping rules can be implemented through sample labeling during model training or user-defined configuration, ultimately forming structured data containing item categories, contour information, and special attributes.

[0042] The above-mentioned optional implementation combines image data and depth information, uses a 3D vision measurement algorithm to accurately identify the outline edge of the object and obtain its size parameters, and uses a pre-trained convolutional neural network model to perform target detection and instance segmentation on the object. This can accurately obtain the object category, outline, occupied space and attribute labels, realizing the automated and high-precision acquisition and analysis of multi-dimensional spatial attribute parameters of the object to be loaded. This provides reliable data support for subsequent trunk space planning, volume judgment and layout strategy generation, effectively avoids errors caused by manual judgment, improves the intelligence level and accuracy of object information acquisition, and thus ensures the rationality and efficiency of subsequent space planning and object placement.

[0043] In some embodiments, when obtaining the first space model of the currently available space in the trunk, the first state parameters of the preset components related to the trunk volume are first obtained, and then the first space model corresponding to the first state parameters is determined from the preset space model library.

[0044] Among them, the pre-set components related to the trunk volume include, but are not limited to: the rear seats of the vehicle and the internal components of the trunk.

[0045] Taking the trunk as an example, the preset components related to its volume include the rear seats, trunk partition / cargo cover, trunk floor, rear center armrest, and sunken trunk storage space. The first state parameters of the preset components include, but are not limited to: the position, angle, or folded-down state of the rear seats, and the internal structure of the trunk. The states of the preset components include, for example, folding down the rear seat backrests, adjusting the rear seat backrest angle, moving the rear seats forward as a whole / in sections, folding / sinking the rear seat cushions, making the rear seats removable / removable, removing or flipping up the trunk partition / cargo cover, lowering the trunk floor / expanding the sunken storage space, and folding / opening up the rear center armrest.

[0046] The status parameters of the components can be obtained through sensors corresponding to the preset components. For example, the position status information of the rear seats can be collected in real time through the seat position sensor. The position signal output by the seat position sensor is collected in real time, and a one-to-one mapping relationship is established between the position signal and the fore-and-aft sliding travel and backrest tilt position of the rear seats.

[0047] The preset space model library is a pre-set library of trunk space models, including the trunk space models corresponding to preset components under different state parameters. The state parameters of the preset components correspond one-to-one with the space models. The preset space model library contains basic space models of the trunk's length, width, and height dimensions, as well as irregular structural locations such as wheel arches and door sill protrusions.

[0048] For example, the preset space model library includes trunk space models associated with the rear seats under different state parameters, such as stepped space models corresponding to the rear seats in the normal seating position, fully folded down, and different fore-and-aft sliding positions.

[0049] The construction process of the pre-set spatial model library is as follows: Physical dimension data of the vehicle's trunk in its factory-delivered state can be collected in advance, including the length, width, and height boundary values ​​of the internal space, as well as the 3D coordinate information of irregular structures such as wheel arches, door sill protrusions, wheel arches, and storage compartments, to construct an initial trunk point cloud model. Based on this point cloud model, a 3D spatial model containing loadable and non-occupiable areas is generated through voxelization or polygon mesh modeling, and stored in the non-volatile memory of the vehicle control unit. Simultaneously, key constraint points such as the trunk opening outline, platform height, and pillar positions are marked in the model, forming the basic spatial model data for subsequent spatial calculations and path planning, providing a benchmark for subsequent dynamic spatial modeling.

[0050] Based on the travel and movement relationship of the vehicle seat adjustment mechanism, a mapping model between the rear seat status and the trunk expansion space is pre-established. First, spatial boundary data of the seat in different adjustment states is collected, including the outer contour coordinates of the backrest and seat cushion when the seat is in the normal seating position, the through space boundary formed when the seat is fully reclined, and the changes in trunk depth corresponding to different positions of the seat's fore-and-aft sliding travel. Then, Boolean operations are performed on the seat contour data of each state with the aforementioned basic spatial model to generate the loadable space envelope of the corresponding state, constructing a normal seating space model, a fully reclined expansion model, and a multi-position stepped space model. Finally, the spatial models of each state are indexed and associated with the signal values ​​of the seat position sensors and stored in the vehicle control unit, providing data support for quickly calling the corresponding spatial model based on the real-time seat status.

[0051] Based on the first state parameters of the preset components, the first space model of the trunk corresponding to the preset space model is matched from the preset space model library, thereby determining the first space corresponding to the first space model (i.e. the current available space of the trunk) and clarifying the actual usable space size and shape of the trunk.

[0052] For example, using the position signal as an index, a pre-stored trunk space model matching the current seat state is called from a preset spatial model library; based on the spatial model, the effective boundary contour and spatial constraints of the trunk in the current state are extracted, and combined with the preset coordinate transformation relationship, a real-time updated three-dimensional voxel model is generated, and finally the actual usable space size and shape of the current trunk are determined.

[0053] The above-mentioned optional implementation method, by pre-constructing a trunk space model library containing different component state parameters, matches the corresponding space model according to the current state parameters of preset components related to trunk volume. This can quickly and accurately determine the real-time space shape of the trunk, improve the efficiency and reliability of space model acquisition, and provide accurate and actual structural basic space data for subsequent assessment of the feasibility of accommodating items, space expansion adjustments, and layout planning. This avoids space planning errors caused by ignoring irregular structures and further ensures the rationality and feasibility of the loading layout.

[0054] In some embodiments, the first spatial model can also be generated by real-time acquisition of point cloud data inside the trunk using sensors such as vehicle-mounted LiDAR and depth cameras, and then generated through point cloud stitching and 3D reconstruction algorithms. Alternatively, it can be constructed by measuring the distance to the trunk boundary and internal obstacles using an onboard ultrasonic sensor array, combined with vehicle body structure parameters. Furthermore, it can be generated by dynamically calculating and generating a corresponding spatial model based on real-time position signals of adjustable components such as rear seats, partitions, and cargo covers, combined with vehicle CAD data, based on the vehicle's Controller Area Network (CAN) bus. In addition, the trunk space shape can be corrected using an inertial measurement unit and vehicle attitude information, and dynamically updated in conjunction with a preset base model to obtain a first spatial model that closely matches the actual situation.

[0055] In some embodiments, after obtaining the spatial attribute parameters of the items to be loaded and the first spatial model of the currently available space in the trunk, an initial placement strategy for the first spatial model is generated based on the spatial attribute parameters. In specific implementation, the first spatial model and spatial attribute parameters are first preprocessed to unify the three-dimensional coordinate system of the first spatial model and all items to be loaded, eliminate outlier values ​​of item size, mark constraint attributes such as items that cannot be inverted, are fragile, and have a center of gravity position, and simplify irregular items into regular bounding box models.

[0056] Subsequently, based on the bounding box model and the first space model, 3D collision detection is performed, and simulated placement is carried out according to reasonable positions; based on the 3D collision detection results and simulation results, it is determined whether the current available space can accommodate the items to be loaded, thus completing the initial space adaptability determination.

[0057] The above embodiments complete the space capacity judgment by planning first and then simulating, without requiring users to actually move and test pack. This reduces manual operation and material damage, and avoids errors caused by relying solely on experience. It makes the capacity judgment results more accurate and reliable, and provides an intuitive basis for subsequent optimization of placement schemes or expansion of storage space, effectively improving the intelligence and convenience of loading planning.

[0058] Optionally, in the process of generating the initial placement strategy of the first spatial model based on spatial attribute parameters, the initial placement strategy for placing the items to be loaded in the first space can be generated according to the first spatial model and spatial attribute parameters, through the embedded 3D bin packing algorithm, genetic algorithm or reinforcement learning model, combined with user preferences.

[0059] Using the current 3D space model of the trunk as the constraint boundary, and combining the 3D dimensions, outlines, categories, and fragile / non-stackable attributes of the items to be loaded, a multi-objective optimization problem is constructed. Then, an embedded 3D packing algorithm, genetic algorithm, or reinforcement learning model is invoked, with the optimization objectives set as maximizing space utilization, optimizing item placement stability, prioritizing fragile item protection, and user preferences. An item layout scheme that satisfies the constraints is generated through iterative solutions. Specifically, the 3D packing algorithm uses heuristic rules for space partitioning and item arrangement, the genetic algorithm searches for the optimal solution space through population iteration and crossover / mutation operations, and the reinforcement learning model outputs layout decisions based on a pre-trained state-action policy network. During the solution process, the placement posture (upright / horizontal / sideways), placement order, and safety gaps between adjacent items are simultaneously verified. Finally, the optimal layout scheme, including the target placement position, posture, placement order, and operation instructions for each item, is output, providing users with directly executable loading guidance.

[0060] Optionally, in generating the initial placement strategy for the first space, the first space model and its spatial attribute parameters are first preprocessed to unify the coordinate system and simplify the items to be loaded into regular bounding box models. Based on the regular bounding box models of the items to be loaded included in the initial placement strategy and the first space model, during the process of unifying the physical coordinate system and performing 3D collision detection and simulated placement, legal placement postures are enumerated for the regular bounding box models, and the corresponding 3D dimension parameters under different placement postures are calculated. The first space model is partitioned, and the placement anchor point positions are determined. The regular bounding box models are sorted according to a preset priority, and trial placements are performed sequentially. Subsequently, the trial placement schemes that pass the 3D collision detection can be evaluated to calculate a comprehensive score, and the initial placement strategy with the optimal comprehensive score can be obtained.

[0061] Specifically, for each item to be loaded, all legal placement postures (standing, horizontal, sideways, etc.) are generated based on its physical properties. The 3D dimensions corresponding to each posture are calculated, and postures that do not meet the loading requirements are eliminated. The first space model is divided into placement areas, and placement anchor point positions are determined based on the space corners and bottom planes, prioritizing positions that fit the boundary and have a lower center of gravity as the initial placement reference. Items are prioritized according to their volume from largest to smallest and their regularity from highest to lowest, and then matched to the layout anchor point positions in sequence. Trial placement is performed in conjunction with the corresponding posture to ensure that there is no interference or overlap between the item and the space boundary or other placed items. Precise collision detection is performed on each trial placement to determine whether the item exceeds the first space range or conflicts with fixed structures such as wheel hubs or seats, eliminating invalid layout schemes.

[0062] Subsequently, effective trial placement schemes were scored, with evaluation dimensions including space utilization, loading center of gravity stability, item placement stability, and ease of retrieval and placement. A weighted comprehensive score was calculated. Optimization strategies such as local search and simulated annealing were used to adjust the item positions, postures, and placement order, generating higher-scoring trial placement schemes through multiple iterations. The trial placement scheme with the highest comprehensive score was selected as the initial placement strategy for the first space, outputting the optimal placement position, posture, and overall placement order for each item.

[0063] The aforementioned optional implementation methods, by enumerating legal placement postures and calculating corresponding size parameters, can fully explore the adaptability of items in space, providing more options for layout optimization. Dividing the space into zones and setting layout anchor points, combined with priority sorting and sequential placement, ensures an orderly and efficient layout process, reducing invalid trial calculations. Introducing 3D collision detection effectively eliminates solutions involving space boundary violations and structural interference, ensuring the generated placement strategy is practically feasible. Furthermore, by comprehensively scoring and iteratively optimizing and selecting the optimal initial placement strategy, space utilization can be improved, making item placement more reasonable and the center of gravity more stable, thereby effectively avoiding space waste and chaotic placement. This provides a precise and reliable strategic foundation for subsequent simulated placement and space capacity judgment, further enhancing the intelligence and practicality of the entire loading planning scheme.

[0064] Optionally, after obtaining the initial placement strategy, the spatial attribute parameters of the items to be loaded are mapped to the physical coordinate system of the first spatial model according to the initial placement strategy to determine the placement area of ​​the items to be loaded; the placement area is compared and verified with the boundary of the first spatial model to determine whether the placement position of the items to be loaded exceeds the constraint range of the first spatial model. If so, it is determined that the first space cannot accommodate the items to be loaded, and then the rear seat adjustment scheme and the corresponding target available space can be generated based on the occupancy status of the rear seats, with the optimization goal of meeting the accommodation requirements of the items to be loaded and maximizing the remaining rear seat space; if not, it is determined that the first space can accommodate the items to be loaded.

[0065] The aforementioned optimal layout scheme of items is mapped onto the current 3D space model of the trunk. Through voxel collision detection or space occupancy simulation, the planned placement area of ​​all items is compared and verified against the effective boundaries and unoccupied areas (such as wheel arches and door sill protrusions) of the space model. This verifies whether the planned position of each item exceeds the spatial constraints or interferes with the fixed structure. Simultaneously, the total volume of all items in the layout scheme is compared with the effective volume of the space model. The space utilization rate is calculated, and it is verified whether the preset constraints such as tailgate closing gap and item safety distance are met. If all items can be accommodated without interference in the current available space and meet all constraints, the adaptation is deemed successful. Otherwise, it is determined that the current available space cannot accommodate all items, triggering the rear seat occupancy status. With the goal of maximizing the accommodation needs of the items to be loaded and the remaining rear seat space, a rear seat adjustment scheme and the corresponding target available space are generated, thus completing the space adaptability verification and providing a reliable judgment basis for subsequent processes.

[0066] The aforementioned optional implementation method achieves a quantitative judgment of the storage capacity by accurately mapping the spatial attribute parameters of the items to the first spatial model to determine the placement area and directly comparing and verifying them with the spatial boundaries. This ensures the accuracy and reliability of the spatial storage capacity results. It provides a clear basis for decision-making regarding whether to expand storage space or adjust component status subsequently.

[0067] In some embodiments, determining whether the current available space in the trunk can accommodate the items to be loaded may further include the following steps: First, determine the target space model required to accommodate the items based on the spatial attribute parameters of the items to be loaded; then, in the physical coordinate system to which the first space model belongs, compare the size relationship between the target space model and the first space model. If the target space model is larger than the first space model, it is determined that the current available space cannot accommodate the items to be loaded.

[0068] Specifically, based on the spatial attribute parameters such as the size, volume, and placement of the item to be loaded, the minimum target space model required to accommodate the item is calculated through modeling. Then, the target space model and the current first space model of the trunk are placed in the same physical coordinate system for quantitative comparison of spatial size, volume, and boundary range. When the target space model is larger than the first space model in overall volume or key dimension, it can be determined that the currently available space is insufficient to accommodate the item to be loaded.

[0069] The above embodiments achieve accurate quantitative comparison of spatial models through a unified coordinate system, which can quickly and objectively complete the spatial capacity judgment, avoiding the errors and subjectivity caused by human experience judgment. At the same time, it provides clear triggering conditions for subsequent spatial expansion and adjustment and layout strategy generation, effectively improving the intelligence level and reliability of vehicle loading space judgment, and laying an accurate data foundation for subsequent optimization of loading schemes.

[0070] After concluding that "the current available space in the trunk cannot accommodate the items to be loaded", based on the occupancy status of the rear seats, and with the optimization goal of maximizing the remaining rear passenger space while meeting the needs of the items to be loaded, a rear seat adjustment scheme and the corresponding target available space are generated.

[0071] The occupancy status of the rear seats includes the occupancy status of each row and each seat (occupancy of people / items, vacancy), the current recline angle of the seat, the fore-and-aft sliding position, the backrest recline angle, the segmented status of split seats, as well as the adjustment travel limits, mechanical limit angles and position ranges of each seat.

[0072] This application proposes to expand storage space by adjusting the rear seats when the current available space in the trunk cannot accommodate the items to be loaded. It should be noted that the occupied seat areas are prohibited from being modified, and the adjustment operation is only performed on the empty seats. The application also aims to achieve two optimization goals: first, the expanded target available space can fully accommodate the items to be loaded; second, after the adjustment, the effective passenger space retained in the rear seats reaches the theoretical maximum value.

[0073] In some embodiments, when performing step S101, based on the occupancy status of the rear seats of the vehicle, a second space model is first determined from the preset space model library when the volume of the maximum remaining rear seating space is greater than or equal to that of the target space model; then, based on the difference between the second state parameters and the first state parameters corresponding to the second space model, a rear seat adjustment scheme is generated, and the target available space corresponding to the second space model is determined.

[0074] Among them, the rear seat adjustment scheme is used to guide users to adjust the first state parameter of the preset component to the second state parameter.

[0075] In the process of determining a second space model from a preset space model library, based on the occupancy status of the rear seats of the vehicle, where the volume of the maximum remaining rear seating space is greater than or equal to that of the target space model, completely vacant and adjustable rear seat units are selected based on the occupancy status of the rear seats, while rear seats that are occupied, have fixed storage, or are mechanically locked are excluded. Combining the vehicle's seat structure (e.g., integral / split, 40 / 60 / 505 layout), adjustable seats are sorted according to the principle of minimizing seating space loss and maximizing storage space gain. Seats furthest from the main seating area, with independent segments, and whose folding / sliding has the least impact on remaining seating space are adjusted first, followed by continuous or multi-connected seat units. Simultaneously, the space expansion and seating space reduction corresponding to different adjustment levels of each seat are marked, establishing a mapping table between adjustment levels and space changes.

[0076] Based on the selected adjustable seats and their mapping relationships, multiple adjustment schemes were iteratively calculated. The second state parameters corresponding to different seat sliding, backrest folding, and seat folding actions were combined in turn, and the additional usable space in the trunk, the overall usable space, the number of remaining rear seats, the effective seating dimensions of the remaining seats, and the integrity of the seating area were calculated for each scheme.

[0077] Subsequently, the target space model required to accommodate the item to be loaded is determined based on the spatial attribute parameters. The minimum target space model that can fully accommodate the item can be calculated through 3D modeling based on the spatial attribute parameters such as the size and volume of the item to be loaded.

[0078] The process begins by identifying a second space model from a pre-defined space model library that maximizes the remaining rear passenger space while maintaining a volume equal to or greater than the target space model. Using the target space model's volume and boundary dimensions as matching criteria, a search is conducted within a pre-established trunk space model library to determine the second space model whose volume is not less than the target space model and maximizes the remaining rear passenger space. The entire process is constrained by two main objectives: first, the total available space must be greater than or equal to the minimum space required to load the items (i.e., the second space model's volume must be greater than or equal to the target space model), eliminating solutions that cannot meet loading requirements; second, among all qualified solutions, the optimal combination is selected based on the area of ​​the remaining passenger space, the number of seats, and the passenger comfort range, to maximize the remaining passenger space. Simultaneously, all adjustment actions are verified to ensure they comply with the seat's mechanical limits, avoiding overtravel and structural interference issues.

[0079] After determining the optimal seat adjustment combination, the specific execution actions are solidified. Precise control parameters such as the target position, backrest angle, folding state, and sliding distance of each adjustable seat are clearly defined. Simultaneously, the target available space corresponding to the second space model is calculated and output, including the three-dimensional dimensions, effective volume, and available placement area of ​​the expanded trunk storage area. The above embodiment achieves rapid and accurate positioning of the trunk expansion space by first determining the space required for the items and then matching suitable spaces from the model library. This eliminates the need for real-time modeling and complex calculations, improving space matching efficiency. It also directly locks in the available expansion space form that meets loading requirements, providing a clear basis for subsequent component state adjustments and item layout planning, effectively avoiding space matching errors, and further improving the rationality and reliability of loading planning. In some embodiments, the rear seat adjustment scheme guides the user to adjust the first state parameters of the preset rear seat components to the second state parameters. The second state parameters (such as seat reclining angle, partition lifting height, cargo cover retraction position, etc.) are combined with the current first state parameters of the preset components to identify the adjustment dimensions, adjustment range, and operation steps between the two, generating a clear and easy-to-understand rear seat adjustment plan. The rear seat adjustment plan clearly marks the preset components to be adjusted in each step, the specific adjustment direction, and the target state. It can also be accompanied by graphic or voice prompts to guide users to gradually adjust the preset components of the rear seats from the first state parameters to the second state parameters, ensuring that the trunk space is expanded from the current usable space to the target usable space.

[0080] In the above embodiments, the rear seat adjustment scheme has clear parameter adjustment guidelines, which can reduce the difficulty for users to adjust the preset components of the rear seats, avoid the inability to effectively expand the available space due to improper operation, and quickly realize the expansion of the trunk. This solves the problems of cumbersome and inaccurate space expansion operations in traditional loading.

[0081] S102. Based on the target available space and the spatial attribute parameters of the items to be loaded, generate a loading layout strategy for the items to be loaded.

[0082] In some embodiments, in the physical coordinate system to which the second spatial model belongs, a loading layout strategy for the items to be loaded is generated based on spatial attribute parameters. The loading layout strategy includes the placement order, placement position, and placement posture of the items to be loaded in the target available space.

[0083] Using the target available space as the new constraint boundary, the 3D bin packing, genetic algorithm or reinforcement learning model is called again. Combined with the changes in the spatial structure after the seat adjustment, the optimal placement, posture and order of each item are replanned to generate a loading layout strategy that adapts to the expanded target available space.

[0084] In practice, the spatial attribute parameters of the items to be loaded are placed in the same physical coordinate system as the second space model. Combined with the logic of 3D spatial layout planning, the priority order of item placement, suitable placement positions (fitting spatial boundaries, avoiding irregular structures), and reasonable placement postures (vertical, horizontal, etc.) are clearly defined. The placement details of each item are refined to form a loading layout strategy that can directly guide user operations. This loading layout strategy provides users with standardized placement guidance, effectively avoiding the blind spots of manual placement, improving item placement efficiency, maximizing the use of the available volume of the second space, and reducing space waste.

[0085] S103, outputs rear seat adjustment schemes and loading layout strategies to guide users to adjust the rear seats and place items to be loaded in the trunk.

[0086] The rear seat adjustment scheme and loading layout strategy work together to form a complete loading guidance process, further improving the intelligence and convenience of vehicle loading.

[0087] The system outputs rear seat adjustment options and loading layout strategies to users through the vehicle's infotainment screen, head-up display (HUD), or voice broadcast system, guiding users to adjust the rear seat position, backrest status, and other operations to expand the space, and to place items in the order presented by the system.

[0088] The loading process is monitored in real time by a camera. Once the placement of the items is confirmed to be consistent with the loading layout strategy, the user is prompted to close the hatch, thus completing the entire intelligent loading process.

[0089] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods according to the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method.

[0090] like Figure 2 As shown, embodiments of this application also provide a trunk space layout device, the device comprising: The adjustment scheme generation module 201 is used to generate a rear seat adjustment scheme and the corresponding target available space based on the occupancy status of the rear seats of the vehicle when the current available space in the trunk of the vehicle cannot accommodate the items to be loaded, with the optimization goal of maximizing the remaining rear seat space to meet the accommodating needs of the items to be loaded. The layout strategy generation module 202 generates a loading layout strategy for the item to be loaded based on the target available space and the spatial attribute parameters of the item to be loaded. Output module 203 outputs the rear seat adjustment scheme and the loading layout strategy to guide the user to adjust the rear seat and place the items to be loaded in the trunk.

[0091] In one specific embodiment, the adjustment scheme generation module 201 is further configured to: obtain the spatial attribute parameters of the items to be loaded, and a first spatial model of the currently available space, wherein the first spatial model is a three-dimensional spatial model of preset components related to the trunk volume in a first state parameter; generate an initial placement strategy for the first spatial model based on the spatial attribute parameters; simulate placing the items to be loaded in the first spatial model according to the initial placement strategy, and determine whether the currently available space corresponding to the first spatial model can accommodate the items to be loaded.

[0092] In one specific embodiment, the adjustment scheme generation module 201 is used to: acquire image data and depth information of the item to be loaded; the depth information is used to characterize the mapping relationship between two-dimensional pixel coordinates and three-dimensional spatial distance in the image data; based on the image data and depth information, identify the item to be loaded to obtain spatial attribute parameters; acquire the first state parameters of the preset component; determine the first spatial model corresponding to the first state parameters from the preset spatial model library; wherein, the preset spatial model library includes the space model of the trunk corresponding to the preset component under different state parameters.

[0093] In one specific embodiment, the spatial attribute parameters include: size information, category, contour information, occupied space, and attribute label; the adjustment scheme generation module 201 identifies the item to be loaded based on image data and depth information to obtain the spatial attribute parameters, which are used for: identifying the contour edge of the item to be loaded based on image data and depth information, to obtain the size information of the item to be loaded; and performing target detection and instance segmentation on the image data based on a pre-trained convolutional neural network model to obtain the category, contour information, occupied space, and attribute label of the item to be loaded.

[0094] In one specific embodiment, the adjustment scheme generation module 201 is used to: unify the physical coordinate system and perform three-dimensional collision detection and simulated placement based on the rule bounding box model and the first space model of the items to be loaded contained in the initial placement strategy; and determine whether the currently available space can accommodate the items to be loaded based on the three-dimensional collision detection results and simulation results.

[0095] In one specific embodiment, the initial placement strategy, which includes the items to be loaded, is used to: determine different placement postures for the regular bounding box model and calculate the corresponding three-dimensional dimension parameters under different placement postures; partition the first space model and determine the placement anchor point positions; sort the regular bounding box model according to a preset priority, and perform trial placement and three-dimensional collision detection sequentially according to the placement anchor point positions.

[0096] In one specific embodiment, the adjustment scheme generation module 201 is used to: map the spatial attribute parameters of the items to be loaded to the physical coordinate system of the first spatial model according to the initial placement strategy, and determine the placement area of ​​the items to be loaded; compare and verify the placement area with the boundary of the first spatial model to determine whether the placement position of the items to be loaded exceeds the constraint range of the first spatial model; if so, determine that the current available space cannot accommodate the items to be loaded; if not, determine that the current available space can accommodate the items to be loaded.

[0097] In one specific embodiment, the adjustment scheme generation module 201 is used to: obtain the spatial attribute parameters of the item to be loaded; determine the target spatial model required to accommodate the item to be loaded based on the spatial attribute parameters; compare the size relationship between the target spatial model and the first spatial model in the physical coordinate system to which the first spatial model belongs; and determine that the currently available space cannot accommodate the item to be loaded if the target spatial model is larger than the first spatial model.

[0098] In one specific embodiment, the adjustment scheme generation module 201 is used to: determine the target space model required to accommodate the items to be loaded based on space attribute parameters when the currently available space cannot accommodate the items to be loaded; determine a second space model from a preset space model library whose volume is greater than or equal to the target space model when the remaining rear seating space is at its maximum, based on the occupancy status of the rear seats of the vehicle; generate a rear seat adjustment scheme based on the difference between the second state parameters and the first state parameters corresponding to the second space model, and determine the target available space corresponding to the second space model; the rear seat adjustment scheme is used to guide the user to adjust the first state parameters of the preset components to the second state parameters.

[0099] In one specific embodiment, the layout strategy generation module 202 is used to: generate a loading layout strategy for the items to be loaded based on spatial attribute parameters in the physical coordinate system to which the second spatial model of the target available space belongs. The loading layout strategy includes the placement order, placement position and placement posture of the items to be loaded in the target available space.

[0100] Regarding the apparatus in the above embodiments, the specific manner in which each unit performs its operation has been described in detail in the embodiments related to the method, and will not be elaborated upon here.

[0101] Figure 3 This is a schematic diagram of the structure of a vehicle provided in an embodiment of this application.

[0102] For example, such as Figure 3As shown, the vehicle includes a memory 301 and a processor 302. The memory 301 stores executable program code 3011, and the processor 302 is used to call and execute the executable program code 3011 to execute the space layout method of the vehicle's trunk.

[0103] This embodiment can divide the vehicle into functional modules based on the above method example. For example, each function can be assigned to a separate module, or two or more functions can be integrated into one processing module. The integrated module can be implemented in hardware. It should be noted that the module division in this embodiment is illustrative and only represents one logical functional division. In actual implementation, there may be other division methods. When dividing each functional module according to its corresponding function, the vehicle may include: an acquisition module, a judgment module, a determination module, a layout strategy generation module, etc. It should be noted that all relevant content of each step involved in the above method embodiment can be referenced from the functional description of the corresponding functional module, and will not be repeated here.

[0104] The vehicle provided in this embodiment is used to execute the above-described method for spatial layout of a vehicle trunk, and thus can achieve the same effect as the above-described implementation method.

[0105] When using integrated units, the vehicle may include a processing module and a storage module. The processing module is used to control and manage the vehicle's actions. The storage module supports the vehicle in executing program code and data.

[0106] The processing module may be a processor or a controller, which can implement or execute various exemplary logic blocks, modules, and circuits as disclosed in this application. The processor may also be a combination of computing functions, such as a combination of one or more microprocessors, a combination of digital signal processing (DSP) and microprocessors, etc., and the storage module may be a memory.

[0107] Embodiments of this application also provide an electronic device, including a memory and a processor, wherein the memory stores a computer program and the processor is configured to run the computer program to perform the steps in any of the above embodiments of the vehicle trunk space layout method.

[0108] Embodiments of this application also provide a computer-readable storage medium storing a computer program, wherein the computer program is configured to execute the steps in any of the above embodiments of the vehicle trunk space layout method when it is run.

[0109] In one exemplary embodiment, the aforementioned computer-readable storage medium may include, but is not limited to, various media capable of storing computer programs, such as a USB flash drive, read-only memory (ROM), random access memory (RAM), portable hard disk, magnetic disk, or optical disk.

[0110] Embodiments of this application also provide a computer program product, which includes a computer program that, when executed by a processor, implements the steps in any of the above embodiments of the vehicle trunk space layout method.

[0111] Embodiments of this application also provide another computer program product, including a non-volatile computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps in any of the above embodiments of the vehicle trunk space layout method.

[0112] The beneficial effects of the above embodiments can be referred to the beneficial effects of the corresponding methods provided above, and will not be repeated here.

[0113] Through the above description of the embodiments, those skilled in the art will understand that, for the sake of convenience and brevity, only the division of the above functional modules is used as an example. In actual applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above.

[0114] In the embodiments provided in this application, it should be understood that the disclosed apparatus and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of modules or units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another device, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0115] In the description of this application, it should be understood that if the terms "upper", "lower", "front", "rear", "left" and "right" are used to indicate the orientation or positional relationship based on the orientation or positional relationship shown in the drawings, they are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the position or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of this application.

[0116] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes the element.

[0117] The above are merely embodiments of this application and are not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of the claims of this application.

Claims

1. A method for spatial layout of a vehicle trunk, characterized in that, include: If the current available space in the vehicle's trunk cannot accommodate the items to be loaded, based on the occupancy status of the vehicle's rear seats, with the optimization goal of maximizing the remaining rear seating space while meeting the occupancy requirements of the items to be loaded; a rear seat adjustment scheme and the corresponding target available space are generated. Based on the target available space and the spatial attribute parameters of the items to be loaded, a loading layout strategy for the items to be loaded is generated; Output the rear seat adjustment scheme and the loading layout strategy to guide the user to adjust the rear seat and place the items to be loaded in the trunk.

2. The method according to claim 1, characterized in that, Before generating a rear seat adjustment scheme and corresponding target available space, based on the occupancy status of the rear seats and with the optimization objective of maximizing the remaining rear seating space while meeting the occupancy requirements of the items to be loaded, when the current available space in the vehicle's trunk cannot accommodate the items to be loaded, the method further includes: Obtain the spatial attribute parameters of the item to be loaded, and the first spatial model of the currently available space. The first spatial model is a three-dimensional spatial model of preset components related to the trunk volume in a first state parameter. Based on the spatial attribute parameters, an initial placement strategy for the first spatial model is generated; According to the initial placement strategy, the items to be loaded are simulated to be placed in the first space model, and it is determined whether the currently available space corresponding to the first space model can accommodate the items to be loaded.

3. The method according to claim 2, characterized in that, The step of obtaining the spatial attribute parameters of the item to be loaded and the first spatial model of the currently available space includes: The image data and depth information of the item to be loaded are obtained; the depth information is used to characterize the mapping relationship between two-dimensional pixel coordinates and three-dimensional spatial distance in the image data. Based on the image data and the depth information, the item to be loaded is identified to obtain the spatial attribute parameters; Obtain the first state parameters of the preset component; The first space model corresponding to the first state parameter is determined from the preset space model library; wherein, the preset space model library includes the space model of the trunk corresponding to the preset component under different state parameters.

4. The method according to claim 3, characterized in that, The spatial attribute parameters include: size information, category, outline information, occupied space, and attribute labels; The step of identifying the object to be loaded based on the image data and the depth information to obtain the spatial attribute parameters includes: Based on the image data and the depth information, a three-dimensional vision measurement algorithm is invoked to identify the outline edge of the item to be loaded, thereby obtaining the size information of the item to be loaded. The image data is used to perform target detection and instance segmentation based on a pre-trained convolutional neural network model to obtain the category, contour information, occupied space and attribute labels of the item to be loaded.

5. The method according to claim 2, characterized in that, The step of simulating the placement of the items to be loaded within the first space model according to the initial placement strategy, and determining whether the currently available space corresponding to the first space model can accommodate the items to be loaded, includes: Based on the bounding box model of the items to be loaded included in the initial placement strategy, and the first spatial model, a unified physical coordinate system is established and three-dimensional collision detection and simulated placement are performed. Based on the 3D collision detection results and simulation results, determine whether the currently available space can accommodate the item to be loaded.

6. The method according to claim 5, characterized in that, The process of unifying the physical coordinate system and performing 3D collision detection and simulated placement based on the bounding box model of the items to be loaded included in the initial placement strategy and the first spatial model, includes: Different placement postures are determined for the regular bounding box model, and the corresponding three-dimensional dimensional parameters under different placement postures are calculated; The first spatial model is divided into partitions, and the positions of the anchor points are determined. The bounding box models are sorted according to a preset priority, and trial placement and 3D collision detection are performed sequentially according to the placement anchor point positions.

7. The method according to claim 2, characterized in that, The step of simulating the placement of the items to be loaded within the first space model according to the initial placement strategy, and determining whether the currently available space corresponding to the first space model can accommodate the items to be loaded, includes: According to the initial placement strategy, the spatial attribute parameters of the items to be loaded are mapped to the physical coordinate system of the first spatial model to determine the placement area of ​​the items to be loaded. The placement area is compared and verified with the boundary of the first space model to determine whether the placement position of the item to be loaded exceeds the constraint range of the first space model. If so, it is determined that the currently available space cannot accommodate the item to be loaded; If not, it is determined that the currently available space is sufficient to accommodate the item to be loaded.

8. The method according to claim 2, characterized in that, After obtaining the spatial attribute parameters of the item to be loaded and the first spatial model of the currently available space, the method further includes: determining the target spatial model required to accommodate the item to be loaded based on the spatial attribute parameters; In the physical coordinate system to which the first spatial model belongs, compare the size relationship between the target spatial model and the first spatial model; If the target space model is larger than the first space model, it is determined that the currently available space cannot accommodate the item to be loaded.

9. The method according to claim 8, characterized in that, When the current available space in the vehicle's trunk cannot accommodate the items to be loaded, based on the occupancy status of the rear seats, and with the optimization objective of maximizing the remaining rear seating space while meeting the accommodating requirements of the items to be loaded, a rear seat adjustment scheme and a corresponding target available space are generated, including: Based on the occupancy status of the rear seats of the vehicle, a second space model is determined from the preset space model library when the volume of the maximum remaining rear seating space is greater than or equal to the target space model. The preset space model library includes the space model of the trunk corresponding to the preset component under different state parameters. Based on the difference between the second state parameter and the first state parameter corresponding to the second space model, the rear seat adjustment scheme is generated, and the target available space corresponding to the second space model is determined; the rear seat adjustment scheme is used to guide the user to adjust the first state parameter of the preset component to the second state parameter.

10. The method according to claim 9, characterized in that, The step of generating a loading layout strategy for the items to be loaded based on the target available space and the spatial attribute parameters of the items to be loaded includes: In the physical coordinate system of the second spatial model of the target available space, a loading layout strategy for the items to be loaded is generated according to the spatial attribute parameters. The loading layout strategy includes the placement order, placement position and placement posture of the items to be loaded in the target available space.

11. A vehicle, characterized in that, include: Memory, used to store computer programs; A processor, configured to implement the steps of the spatial layout method for the vehicle trunk as described in any one of claims 1 to 10 when executing the computer program.

Images