A cockpit wallpaper generation method and device, electronic equipment and storage medium

Abstract

Embodiments of the present application provide a cockpit wallpaper generation method and device, electronic equipment and storage medium, the method comprising: acquiring an initial image collected by an image acquisition device of a target vehicle for an environment in which the target vehicle is located; determining orientation information and attribute information of a target moving object based on the initial image; generating a target prompt word corresponding to a target image generation model according to the orientation information and the attribute information, the target image generation model being used to generate a cockpit wallpaper of a preset style based on a prompt word and an image; inputting the target prompt word and the initial image into the target image generation model, and outputting a cockpit wallpaper image of the target vehicle. Through the embodiments of the present application, intelligent and more flexible personalized wallpaper is generated by analyzing the moving object in combination with the image generation model.

Description

Technical Field

[0001] This invention relates to the field of vehicle intelligent cockpit technology, and in particular to a method and apparatus for generating cockpit wallpaper, electronic equipment, and storage medium. Background Technology

[0002] Currently, users can set wallpapers on the in-vehicle infotainment system according to their needs. However, the wallpapers on the in-vehicle infotainment system are usually pictures that users have pre-stored on the system, resulting in a limited number of wallpapers and a lack of novelty and appeal due to the monotony and fixed nature of the displayed wallpapers. Summary of the Invention

[0003] In view of the above problems, a method and apparatus for generating cabin wallpaper, an electronic device, and a storage medium are proposed to overcome or at least partially solve the above problems, including:

[0004] A method for generating cockpit wallpaper, the method comprising:

[0005] Acquire initial images of the target vehicle captured by the image acquisition device for the environment in which the target vehicle is located;

[0006] Based on the initial image, determine the orientation and attribute information of the target moving object;

[0007] Based on the orientation information and the attribute information, a target prompt word is generated corresponding to the target image generation model. The target image generation model is used to generate a cockpit wallpaper with a preset style based on the prompt word and the image.

[0008] The target prompt and the initial image are input into the target image model, and the cockpit wallpaper image of the target vehicle is output.

[0009] Optionally, determining the orientation and attribute information of the target moving object based on the initial image includes:

[0010] The initial image is input into a motion object detection model for detecting moving objects, and the target bounding box information of the target moving object is output.

[0011] The target bounding box information is input into an attribute recognition model used to extract the attribute information of a moving object, and the attribute information of the target moving object is output.

[0012] The target bounding box information is input into a depth estimation model used to estimate the depth information of a moving object, and the depth information from the center point of the target bounding box of the moving object to the center point of the viewpoint of the image acquisition device is output.

[0013] The orientation information of the moving target object is determined based on the target bounding box information and the depth information.

[0014] Optionally, determining the orientation information of the moving target object based on the target bounding box information and the depth information includes:

[0015] Determine whether the depth information is less than a preset depth threshold;

[0016] When it is determined that the depth information is less than the preset depth threshold, a first vector from the center point of the view of the image acquisition device to the center point of the target box information is determined, and a second vector from the center point of the view of the image acquisition device to the coordinates of the lower right corner of the initial image is determined;

[0017] Determine the target angle between the first vector and the second vector;

[0018] The orientation information of the moving target is determined based on the target angle.

[0019] Optionally, generating target prompt words corresponding to the target image model based on the location information and the attribute information includes:

[0020] The orientation information and attribute information are preprocessed to obtain the structured information corresponding to the target image model;

[0021] Extract multiple key feature information from the structured information;

[0022] The multiple key features are integrated to obtain integrated feature information;

[0023] The integrated feature information is expanded to obtain the target feature information;

[0024] Based on target feature information, target prompt words are generated corresponding to the target image model.

[0025] Optionally, the method further includes:

[0026] Obtain scene information about the environment in which the target vehicle is located;

[0027] The step of generating target prompt words corresponding to the target image model based on the orientation information and the attribute information includes:

[0028] Based on the location information, attribute information, and scene information, target prompt words are generated corresponding to the target image model.

[0029] Optionally, obtaining the scene information of the environment in which the target vehicle is located includes:

[0030] Obtain the real-time location information of the target vehicle;

[0031] The scene information of the target vehicle's environment is determined based on the real-time location information.

[0032] Optionally, the initial image acquired by the image acquisition device for the target vehicle, based on the environment in which the target vehicle is located, includes:

[0033] When a preset gesture of a target user inside the target vehicle is detected, the wallpaper generation function is activated;

[0034] When the wallpaper generation function is enabled, the initial image of the target vehicle is acquired by the image acquisition device for the environment in which the target vehicle is located.

[0035] Optionally, the method further includes:

[0036] The cabin wallpaper image is displayed on the in-vehicle infotainment system and / or ceiling screen of the target vehicle.

[0037] An apparatus for generating cabin wallpaper, the apparatus comprising:

[0038] An initial image acquisition module is used to acquire an initial image of the target vehicle captured by the image acquisition device of the target vehicle in relation to the environment in which the target vehicle is located;

[0039] The information extraction module is used to determine the orientation and attribute information of the target moving object based on the initial image;

[0040] The prompt word generation module is used to generate target prompt words corresponding to the target image generation model based on the orientation information and the attribute information. The target image generation model is used to generate cockpit wallpapers of a preset style based on the prompt words and images.

[0041] The cockpit wallpaper generation module is used to input the target prompt word and the initial image into the target image model and output the cockpit wallpaper image of the target vehicle.

[0042] An electronic device includes a processor, a memory, and a computer program stored in the memory and capable of running on the processor, wherein the computer program, when executed by the processor, implements the cockpit wallpaper generation method described above.

[0043] A computer-readable storage medium storing a computer program that, when executed by a processor, implements the cockpit wallpaper generation method described above.

[0044] The embodiments of the present invention have the following advantages:

[0045] In this embodiment of the invention, by acquiring an initial image of the target vehicle's environment using an image acquisition device, the orientation and attribute information of the moving target can be determined based on the initial image. Then, target prompts corresponding to the target image-generated model are generated based on the orientation and attribute information. Furthermore, the target prompts and the initial image are input into the target image-generated model, thereby outputting a cabin wallpaper image of the target vehicle. This achieves intelligent and more flexible personalized wallpaper generation by analyzing the moving object and combining it with the image-generated model. Attached Figure Description

[0046] To more clearly illustrate the technical solution of the present invention, the accompanying drawings used in the description of the present invention will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0047] Figure 1a This is a schematic diagram of a wallpaper generation process in the existing technology;

[0048] Figure 1b This is a flowchart of a method for generating cabin wallpaper according to an embodiment of the present invention;

[0049] Figure 2 This is a flowchart of another method for generating cabin wallpaper according to an embodiment of the present invention;

[0050] Figure 3a This is a schematic diagram of a cabin wallpaper generation process provided by an embodiment of the present invention.

[0051] Figure 3b This is an initial image captured by an ADAS camera according to an embodiment of the present invention;

[0052] Figure 3c This invention provides a cartoon-style wallpaper image generated based on a wallpaper generation function, as provided in one embodiment of the present invention.

[0053] Figure 4 This is a schematic diagram of the structure of a cabin wallpaper generation device provided in an embodiment of the present invention. Detailed Implementation

[0054] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.

[0055] Currently, the wallpapers displayed on the infotainment systems of mainstream car manufacturers are generally pre-stored by users. While switching wallpapers using these stored wallpapers provides a fast and stable user experience, its inherent disadvantages cannot be ignored. The number and variety of predefined wallpapers in the current settings are limited, failing to meet the increasingly diverse and personalized needs of users. As users' demands for the aesthetics and personalization of the in-car environment increase, single and fixed wallpaper options appear too restrictive, lacking novelty and appeal. Furthermore, although predefined wallpapers allow for quick switching, this also means that these wallpapers lack dynamic changes and the possibility of personalized customization; users can only choose from a limited selection, resulting in a relatively monotonous experience. To overcome these disadvantages, a more intelligent and flexible method for generating personalized wallpapers is urgently needed.

[0056] A flowchart illustrating the traditional predefined wallpaper switching method is shown below. Figure 1a The steps shown are as follows:

[0057] (1) Upload wallpaper: Vehicle users can choose one or more wallpapers to be pre-installed on the vehicle system according to their own preferences.

[0058] (2) Voice input: The vehicle system receives the voice command "Help me change the wallpaper" from the driver and passenger.

[0059] (3) Speech signal processing: Process speech signals to reduce noise and echo.

[0060] (4) Voice recognition: Recognize speech and convert it into text "Help me change the wallpaper".

[0061] (5) Natural Language Processing: Understanding the user's intent is to help me change the wallpaper.

[0062] (6) Dialogue Management: Confirm the current wallpaper settings and decide whether to execute the command to switch wallpapers.

[0063] (7) Execute control command: When it is determined that the wallpaper switching command can be executed, you can send "Switch wallpaper".

[0064] (8) Feedback Confirmation: Confirm that the wallpaper has been successfully switched and generate feedback information.

[0065] This wallpaper setting method lacks personalized wallpaper generation services and only meets the interaction needs of the driver and front passenger, while the experience of rear passengers is often neglected, and they cannot enjoy the same convenience and personalized services.

[0066] In this embodiment of the invention, an initial image of the target vehicle is acquired by an image acquisition device for the environment in which the target vehicle is located; the orientation and attribute information of the target moving object are determined based on the initial image; target prompt words are generated according to the orientation and attribute information, and the target image model is used to generate a cabin wallpaper of a preset style based on the prompt words and the image; the target prompt words and the initial image are input into the target image model, and the cabin wallpaper image of the target vehicle is output, thereby enabling the generation of intelligent and more flexible personalized wallpapers by analyzing moving objects and combining them with the image model.

[0067] Reference Figure 1b The diagram illustrates a flowchart of a method for generating cockpit wallpaper according to an embodiment of the present invention, which may specifically include the following steps:

[0068] Step S101: Acquire the initial image of the target vehicle captured by the image acquisition device for the environment in which the target vehicle is located;

[0069] In practical applications, the target vehicle can be equipped with an image acquisition device. This device can be used to capture initial images of the interior and exterior of the vehicle to assist in better driving. The image acquisition device can be a camera; specifically, in this embodiment of the invention, it can be an ADAS (Advanced Driver Assistance Systems) camera. In this embodiment, by calling the ADAS camera to capture the scene ahead, a real-time environmental image of the target vehicle is obtained, and then a wallpaper effect related to the actual driving environment can be generated based on this image.

[0070] In one embodiment of the present invention, a wallpaper generation function can be set. Users can use a variety of different preset activation methods to activate the wallpaper generation function. After activating the wallpaper generation function, step S101 is executed to obtain an initial image, and then a wallpaper image is generated based on the initial image.

[0071] In one example, when a preset gesture from a target user inside the target vehicle is detected, the wallpaper generation function is activated. This allows the acquisition of an initial image of the target vehicle's environment by its image acquisition device, while the wallpaper generation function is enabled. The preset gesture can be set according to the actual scenario; in this embodiment of the invention, no excessive restrictions are placed on it. For example, the preset gesture could be a dynamic gesture used to click the function button for generating the cockpit wallpaper.

[0072] Specifically, an image acquisition device can be used to capture user images corresponding to the target user. These images are then analyzed to determine if a preset gesture of the target user is detected. If the preset gesture is detected, the wallpaper generation function can be activated. If no preset gesture is detected, the next period's user images are acquired and analyzed.

[0073] In this embodiment of the invention, the wallpaper generation function is triggered by air gestures, achieving the effect of dynamically generating personalized wallpapers in the cabin without the user having to directly touch the screen.

[0074] In another example, the target vehicle can be set to use voice interaction, and the target user can then activate the wallpaper generation function through voice interaction with the target vehicle.

[0075] In another example, the target vehicle's central control screen can be a touch screen, allowing users to directly activate the wallpaper generation function through touch operations on the touch screen.

[0076] Step S102: Determine the orientation and attribute information of the target moving object based on the initial image;

[0077] After acquiring the initial image, image analysis processing can be performed on the initial image to determine the orientation and attribute information of the target moving object present in the initial image. The attribute information can include the target moving object's own attributes and the target moving object's motion attributes. The specific type of attribute information can be determined according to the type of the target moving object.

[0078] Taking pedestrians as the target user group as an example, their own attributes may include, but are not limited to, the pedestrian's gender and age; their movement attributes may include, but are not limited to, whether they have a vehicle and the type of vehicle, such as the pedestrian riding a motorcycle or a bicycle.

[0079] Step S103: Generate target prompt words corresponding to the target image generation model based on the orientation information and attribute information. The target image generation model is used to generate cockpit wallpapers with preset styles based on the prompt words and images.

[0080] After obtaining the location and attribute information, target prompts in a preset format can be generated based on this information. This preset format represents the format of prompts that can be input into the target graph-to-graph model. The preset format can be determined based on the actual graph-to-graph model obtained through training.

[0081] In this embodiment of the invention, the input data of the target image model is a prompt word and an initial image, and the output data is a cockpit wallpaper with a preset style. The preset style can be set according to actual needs. For example, the preset style in this embodiment of the invention can be a comic style.

[0082] In the target image generation model, the input prompts are in a preset format. The initial image can be transformed according to a preset style. During the transformation process, the prompts further enhance the features of the target moving object so that a target moving object of the preset style can be generated when the initial image is transformed according to the preset style.

[0083] In one embodiment of the present invention, generating target prompt words corresponding to the target image model based on location information and attribute information may specifically include the following sub-steps:

[0084] Sub-step S11 involves preprocessing the orientation and attribute information to obtain the structured information corresponding to the target image generation model.

[0085] Location and attribute information are key data for describing moving objects and scenes. To enable this information to be effectively utilized by target graph models (such as convolutional neural networks and graph neural networks), data preprocessing is required to transform it into structured information usable by the target graph model. Preprocessing can specifically include one or more of the following: data collection and alignment, and data standardization.

[0086] In practical applications, orientation and attribute information can be standardized to meet the input requirements of the target image generation model. Data standardization can specifically include, but is not limited to, coordinate standardization, velocity and direction standardization, and attribute standardization.

[0087] Sub-step S12 extracts multiple key feature information from the structured information;

[0088] After obtaining structured information, key features can be selected from it, such as object type, position, speed, direction, color, and size. Then, the selected key features are extracted from the structured information of each moving target object.

[0089] Sub-step S13 integrates multiple key features to obtain integrated feature information;

[0090] The key features of each object are concatenated into a feature vector.

[0091] Sub-step S14 involves expanding the integrated feature information to obtain the target feature information;

[0092] Based on the integrated feature vectors, a natural language description is generated. Furthermore, contextual information can be added to expand the description and obtain more comprehensive target feature information.

[0093] Sub-step S15: Generate target prompt words corresponding to the target image model based on target feature information.

[0094] Transform natural language descriptions into a format that the target graph model can understand, such as structured data formats or specific prompt templates, i.e., target prompt words.

[0095] In one embodiment of the present invention, scene information of the environment in which the target vehicle is located can be obtained in advance; then, target prompt words corresponding to the target image model can be generated based on the orientation information, attribute information and scene information.

[0096] In practical applications, location information, attribute information, and scene information can be combined for preprocessing, integration, and expansion to generate target prompt words corresponding to the target image model, so as to generate wallpapers that fit the current vehicle's environment.

[0097] In this embodiment of the invention, by integrating scene information and moving object analysis information (i.e., orientation information and attribute information), a structured input is provided for the graph-to-graph model, making the generated wallpaper more in line with the atmosphere of the current scene.

[0098] In one example, scene information may include weather conditions (sunny, rainy, sunny, etc.) at the location of the user's vehicle and time of day (morning, afternoon, and evening).

[0099] In this embodiment of the invention, the real-time location information of the target vehicle can be obtained; then, based on the real-time location information, the scene information of the environment in which the target vehicle is located can be determined through a preset interface. For example, weather information and clock information can be obtained by calling the preset interface based on the real-time location information, thereby determining the weather status and the current time period at the real-time location.

[0100] Step S104: Input the target prompt and the initial image into the target image model, and output the cockpit wallpaper image of the target vehicle.

[0101] After obtaining the target prompt, the target prompt and the initial image can be input into the trained target graph-generated image model. The graph-generated image model can then generate a cockpit wallpaper image in a preset style. In practical applications, the graph-generated image model can be trained according to various wallpaper styles, and can then generate a cockpit wallpaper image corresponding to the initial image based on the user's selected style or the user's historical wallpaper styles.

[0102] For example, graph-generated graph models can be trained based on diffusion models, which include, but are not limited to, models such as SDXL (Stable Diffusion XL) and DIL (Deep Integration Learning Model).

[0103] In this embodiment of the invention, different styles of wallpapers can be generated through the image generation module, achieving the effect of coordinating the cabin wallpaper with the actual scene atmosphere, enhancing the user's immersion and aesthetic experience.

[0104] In one embodiment of the present invention, a cabin wallpaper image is displayed on the in-vehicle infotainment system and / or ceiling-mounted screen of the target vehicle. By mapping the generated wallpaper onto the in-vehicle infotainment system or ceiling-mounted screen, the effect of instantly displaying personalized wallpapers can be achieved, enhancing the user experience and the technological feel of the cabin.

[0105] In this embodiment of the invention, by acquiring an initial image of the target vehicle's environment using an image acquisition device, the orientation and attribute information of the moving target can be determined based on the initial image. Then, target prompts corresponding to the target image-generated model are generated based on the orientation and attribute information. Furthermore, the target prompts and the initial image are input into the target image-generated model, thereby outputting a cabin wallpaper image of the target vehicle. This achieves intelligent and more flexible personalized wallpaper generation by analyzing the moving object and combining it with the image-generated model.

[0106] Reference Figure 2 The diagram illustrates a flowchart of another method for generating cockpit wallpaper according to an embodiment of the present invention, which may specifically include the following steps:

[0107] Step S201: Acquire the initial image of the target vehicle captured by the image acquisition device for the environment in which the target vehicle is located;

[0108] Step S202: Input the initial image into the motion object detection model for detecting moving objects, and output the target bounding box information of the target moving object;

[0109] After obtaining the initial image, it can be preprocessed, such as image scaling, image normalization, and image enhancement.

[0110] The preprocessed image is then input into the trained moving object detection model, which outputs bounding box information, including the object's category, confidence level, and position (x, y, width, height).

[0111] In this embodiment of the invention, target detection can be achieved using the YOLO (You Only Look Once) or SSD (SingleShot MultiBox Detector) models.

[0112] Step S203: Input the target bounding box information into the attribute recognition model used to extract the attribute information of the moving object, and output the attribute information of the target moving object;

[0113] Then, the target bounding box information and the initial image can be input into the attribute recognition model. The attribute recognition model can perform attribute recognition on the moving target object within the target bounding box in the initial image, thus identifying the attribute information of the moving target object.

[0114] Specifically, attribute recognition models can be trained using convolutional neural networks (CNNs) or residual networks (ResNets) to achieve attribute recognition.

[0115] Step S204: Input the target bounding box information into the depth estimation model used to estimate the depth information of the moving object, and output the depth information from the center point of the target bounding box of the moving object to the center point of the view of the image acquisition device.

[0116] After obtaining the target bounding box information, depth estimation can be performed in the depth estimation model by combining the target bounding box information and the initial image to determine the depth information from the center point of the target bounding box of the moving object to the center point of the viewpoint of the image acquisition device.

[0117] For example, depth estimation can be performed using monocular depth estimation models (such as Monodepth2) or binocular depth estimation models (such as PSMNet).

[0118] In this embodiment of the invention, some edge-side multimodal models with fewer parameters can also be used to replace the moving object detection model, attribute recognition model and depth estimation model, so as to output the orientation information and attribute information obtained in the initial image by inputting the initial image to the edge-side multimodal model.

[0119] Step S205: Determine the orientation information of the moving target object based on the target bounding box information and depth information.

[0120] After obtaining the target bounding box information and depth information, the specific location of the moving target object can be further determined.

[0121] In one embodiment of the present invention, determining the orientation information of a moving target object based on the target bounding box information and depth information includes the following sub-steps:

[0122] Sub-step S21: Determine whether the depth information is less than a preset depth threshold;

[0123] Since moving objects appear blurry when their size in the frame is too small, they can be ignored. In this embodiment of the invention, a preset depth threshold can be set to filter valid target moving objects. The preset depth threshold can be set according to the actual scene, and this embodiment of the invention does not impose many restrictions on it.

[0124] Sub-step S22: When the depth information is determined to be less than the preset depth threshold, a first vector from the center point of the view of the image acquisition device to the center point of the target box information and a second vector from the center point of the view of the image acquisition device to the coordinates of the lower right corner of the initial image are determined.

[0125] When the depth information is less than the preset depth threshold, it is necessary to determine the orientation of the moving target. The specific orientation can be determined according to the view coordinate system under the view of the image acquisition device, such as determining the coordinates of the view center point and the center point of the target box information, and then determining the first vector from the view center point to the center point of the target box information; determining the coordinates of the view center point of the image acquisition device and the coordinates of the lower right corner of the initial image, and thus determining the second vector from the view center point of the image acquisition device to the coordinates of the lower right corner of the initial image.

[0126] Sub-step S23: Determine the target angle between the first vector and the second vector;

[0127] The angle between vectors can be calculated from the two vectors. The specific calculation process is as follows:

[0128]

[0129] Where θ represents the target angle, OA0 represents the second vector, and OA1 represents the first vector.

[0130] Sub-step S24: Determine the orientation information of the moving target based on the target angle.

[0131] After calculating the target angle, the azimuth information can be determined based on the target angle. For example, when θ is between 75° and 105°, it is determined to be directly in front; when θ is between 0° and 75°, it is determined to be to the right front; and when θ is between 105° and 180°, it is determined to be to the left front.

[0132] Step S206: Generate target prompt words corresponding to the target image generation model based on the orientation information and attribute information. The target image generation model is used to generate cockpit wallpapers with preset styles based on the prompt words and images.

[0133] Step S207: Input the target prompt word and the initial image into the target image model, and output the cockpit wallpaper image of the target vehicle.

[0134] This invention combines a motion object detection model, attribute recognition, and depth estimation model to accurately analyze moving objects in the foreground scene and ensure that the generated wallpaper contains important visual objects, such as people and animals.

[0135] Reference Figure 3a The diagram shown illustrates the generation process of a cabin wallpaper according to an embodiment of the present invention, which may include the following steps:

[0136] (1) Air gesture triggers wallpaper generation function: Use dynamic gestures to click the function button for generating cabin wallpaper to trigger the wallpaper generation function of the cabin ceiling screen or the front vehicle screen;

[0137] (2) Activate ADAS camera: At this time, ADAS performs a photo capture function on the scene in front to obtain an initial image, such as... Figure 3b As shown.

[0138] (3) Moving object analysis: Using pre-trained moving object detection models, attribute recognition models and depth estimation models, the location information of moving objects (such as the target box information in the picture and whether the position is directly in front of the car, to the right or to the left) and related motion attributes (such as pedestrians riding motorcycles, pedestrians riding bicycles, etc.) and the approximate age and gender attributes of pedestrians are determined. The relevant location information and moving object related attribute information are output in a structured form.

[0139] The general algorithm flow for determining the location of moving objects in the moving object analysis module is as follows:

[0140] The system uses a detection model to extract bounding boxes for moving objects and a depth estimation model to estimate the depth of frame images. By calculating the center point of the moving object's bounding box, the distance between the moving object and the vehicle is determined based on the depth value from the center point to the center point in the camera's view. A depth threshold K is set to determine whether the moving object is a usable scene object. If the moving object is usable, its location is further determined.

[0141] To determine whether a moving object is directly in front, to the right, or to the left, we can define vectors OA1, OA2, OA3, ..., O between the bottom center point of the image and the center point of the moving object's bounding box, and OA0 between the bottom center point of the image and the bottom right corner. We can then calculate the angle between these vectors, using OA1 and OA0 as an example, as shown in the following formula:

[0142]

[0143] When θ is between 75° and 105°, it is considered directly in front; when θ is between 0° and 75°, it is considered to be to the right; and when θ is between 105° and 180°, it is considered to be to the left.

[0144] (4) Forward Scene Analysis: This step uses some interfaces to obtain the weather conditions (sunny, rainy, sunny, etc.) and time period information (morning, afternoon, and evening) of the user's vehicle location.

[0145] (5) Prompt expansion module: This module integrates and expands the structured information output by the motion object analysis module and the structured information output by the scene analysis module, and transforms it into the prompts required for model input;

[0146] (6) Image Generation Module: The input to this module is prompts that have been expanded and integrated by the Prompt expansion module, and corresponding scene images, such as... Figure 3b As shown, the output is a comic-style image that corresponds to the atmosphere of the scene image (a "rainy day" was added to demonstrate the effect; in reality, the atmosphere of the two should complement each other). Figure 3c As shown.

[0147] It should be noted that, for the sake of simplicity, the method embodiments are described as a series of actions. However, those skilled in the art should understand that the embodiments of the present invention are not limited to the described order of actions, because according to the embodiments of the present invention, some steps can be performed in other orders or simultaneously. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are preferred embodiments, and the actions involved are not necessarily essential to the embodiments of the present invention.

[0148] Reference Figure 4 The diagram shows a structural schematic of a cabin wallpaper generation device according to an embodiment of the present invention, which may specifically include the following modules:

[0149] The initial image acquisition module 401 is used to acquire the initial image of the target vehicle captured by the image acquisition device for the environment in which the target vehicle is located;

[0150] Information extraction module 402 is used to determine the orientation information and attribute information of the target moving object based on the initial image;

[0151] The prompt word generation module 403 is used to generate target prompt words corresponding to the target image generation model based on the orientation information and the attribute information. The target image generation model is used to generate cockpit wallpapers of a preset style based on the prompt words and images.

[0152] The cockpit wallpaper generation module 404 is used to input the target prompt word and the initial image into the target image model and output the cockpit wallpaper image of the target vehicle.

[0153] In one embodiment of the present invention, the information extraction module 402 may include:

[0154] The target bounding box information determination submodule is used to input the initial image into a motion object detection model for detecting moving objects and output the target bounding box information of the target moving object.

[0155] The attribute information determination submodule is used to input the target bounding box information into the attribute recognition model for extracting the attribute information of the moving object, and output the attribute information of the target moving object;

[0156] The depth estimation submodule is used to input the target bounding box information into a depth estimation model for estimating the depth information of a moving object, and output the depth information from the center point of the target bounding box of the moving object to the center point of the viewpoint of the image acquisition device.

[0157] The orientation information determination submodule is used to determine the orientation information of the target moving object based on the target bounding box information and the depth information.

[0158] In one embodiment of the present invention, the orientation information determination submodule may include the following units:

[0159] A depth determination unit is used to determine whether the depth information is less than a preset depth threshold.

[0160] The vector determination unit is used to determine, when it is determined that the depth information is less than the preset depth threshold, a first vector from the center point of the viewpoint of the image acquisition device to the center point of the target box information, and a second vector from the center point of the viewpoint of the image acquisition device to the coordinates of the lower right corner of the initial image;

[0161] A vector angle calculation unit is used to determine the target angle between the first vector and the second vector;

[0162] The orientation determination unit is used to determine the orientation information of the moving target object based on the target angle.

[0163] In one embodiment of the present invention, the prompt word generation module 403 may include the following sub-modules:

[0164] The preprocessing submodule is used to preprocess the orientation information and the attribute information to obtain the structured information corresponding to the target image model;

[0165] The key information extraction submodule is used to extract multiple key feature information from the structured information;

[0166] The integration processing submodule is used to integrate the multiple key features to obtain integrated feature information;

[0167] An expansion processing submodule is used to expand the integrated feature information to obtain target feature information;

[0168] The target prompt word generation submodule is used to generate target prompt words corresponding to the target image model based on target feature information.

[0169] In one embodiment of the present invention, the device further includes:

[0170] The scene information acquisition module is used to acquire scene information of the environment in which the target vehicle is located;

[0171] When the prompt word generation module 403 generates target prompt words corresponding to the target image model based on the location information and the attribute information, it is specifically used for:

[0172] Based on the location information, attribute information, and scene information, target prompt words are generated corresponding to the target image model.

[0173] In one embodiment of the present invention, the scene information acquisition module includes:

[0174] The real-time location acquisition submodule is used to acquire the real-time location information of the target vehicle;

[0175] The scene information determination submodule is used to determine the scene information of the environment in which the target vehicle is located based on the real-time location information.

[0176] In one embodiment of the present invention, the initial image acquisition module 401 may include the following sub-modules:

[0177] The wallpaper generation function enable submodule is used to enable the wallpaper generation function when a preset gesture of the target user inside the target vehicle is detected;

[0178] The initial image acquisition submodule is used to acquire the initial image of the target vehicle captured by the image acquisition device for the environment in which the target vehicle is located when the wallpaper generation function is enabled.

[0179] In one embodiment of the present invention, the device further includes:

[0180] The wallpaper display module is used to display the cabin wallpaper image on the in-vehicle infotainment system and / or ceiling screen of the target vehicle.

[0181] In this embodiment of the invention, by acquiring an initial image of the target vehicle's environment using an image acquisition device, the orientation and attribute information of the moving target can be determined based on the initial image. Then, target prompts corresponding to the target image-generated model are generated based on the orientation and attribute information. Furthermore, the target prompts and the initial image are input into the target image-generated model, thereby outputting a cabin wallpaper image of the target vehicle. This achieves intelligent and more flexible personalized wallpaper generation by analyzing the moving object and combining it with the image-generated model.

[0182] An embodiment of the present invention also provides an electronic device, which may include a processor, a memory, and a computer program stored in the memory and capable of running on the processor. When the computer program is executed by the processor, it implements the above method for generating cockpit wallpaper.

[0183] An embodiment of the present invention also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the above-described method for generating cockpit wallpaper.

[0184] As the device embodiment is basically similar to the method embodiment, the description is relatively simple, and relevant parts can be found in the description of the method embodiment.

[0185] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. The same or similar parts between the various embodiments can be referred to each other.

[0186] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, apparatus, or computer program products. Therefore, embodiments of the present invention can take the form of entirely hardware embodiments, entirely software embodiments, or embodiments combining software and hardware aspects. Furthermore, embodiments of the present invention can take the form of computer program products implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0187] Embodiments of the present invention are described with reference to flowchart illustrations and / or block diagrams of methods, terminal devices (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing terminal device to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing terminal device, create means for implementing the functions specified in one or more blocks of the flowchart illustrations and / or one or more blocks of the block diagrams.

[0188] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing terminal device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means that implement the functions specified in one or more flowcharts and / or one or more block diagrams.

[0189] These computer program instructions may also be loaded onto a computer or other programmable data processing terminal equipment to cause a series of operational steps to be performed on the computer or other programmable terminal equipment to produce a computer-implemented process, such that the instructions, which execute on the computer or other programmable terminal equipment, provide steps for implementing the functions specified in one or more flowcharts and / or one or more block diagrams.

[0190] Although preferred embodiments of the present invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of the embodiments of the present invention.

[0191] Finally, 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. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or terminal device that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or terminal device. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or terminal device that includes said element.

[0192] The above provides a detailed description of the method, apparatus, electronic device, and storage medium for generating cockpit wallpaper. Specific examples have been used to illustrate the principles and implementation methods of the present invention. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of the present invention. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of the present invention. Therefore, the content of this specification should not be construed as a limitation of the present invention.

Claims

1. A method for generating cockpit wallpaper, characterized in that, The method includes: Acquire initial images of the target vehicle captured by the image acquisition device for the environment in which the target vehicle is located; Based on the initial image, determine the orientation and attribute information of the target moving object; the attribute information includes the target moving object's own attributes and the target moving object's motion attributes; Based on the orientation information and the attribute information, a target prompt word is generated corresponding to the target image generation model. The target image generation model is used to generate a cockpit wallpaper with a preset style based on the prompt word and the image. The target prompt and the initial image are input into the target image model, and the cockpit wallpaper image of the target vehicle is output. The determination of the orientation and attribute information of the target moving object based on the initial image includes: The initial image is input into a motion object detection model for detecting moving objects, and the target bounding box information of the target moving object is output. The target bounding box information is input into an attribute recognition model used to extract the attribute information of a moving object, and the attribute information of the target moving object is output. The target bounding box information is input into a depth estimation model used to estimate the depth information of a moving object, and the depth information from the center point of the target bounding box of the moving object to the center point of the viewpoint of the image acquisition device is output. The orientation information of the moving target object is determined based on the target bounding box information and the depth information.

2. The method according to claim 1, characterized in that, Determining the orientation information of the moving target object based on the target bounding box information and the depth information includes: Determine whether the depth information is less than a preset depth threshold; When it is determined that the depth information is less than the preset depth threshold, a first vector from the center point of the view of the image acquisition device to the center point of the target box information is determined, and a second vector from the center point of the view of the image acquisition device to the coordinates of the lower right corner of the initial image is determined; Determine the target angle between the first vector and the second vector; The orientation information of the moving target is determined based on the target angle.

3. The method according to claim 1, characterized in that, The step of generating target prompt words corresponding to the target image model based on the orientation information and the attribute information includes: The orientation information and attribute information are preprocessed to obtain the structured information corresponding to the target image model; Extract multiple key feature information from the structured information; The multiple key features are integrated to obtain integrated feature information; The integrated feature information is expanded to obtain the target feature information; Based on target feature information, target prompt words are generated corresponding to the target image model.

4. The method according to any one of claims 1 to 3, characterized in that, The method further includes: Obtain scene information about the environment in which the target vehicle is located; The step of generating target prompt words corresponding to the target image model based on the orientation information and the attribute information includes: Based on the location information, attribute information, and scene information, target prompt words are generated corresponding to the target image model.

5. The method according to claim 4, characterized in that, The step of obtaining scene information about the environment in which the target vehicle is located includes: Obtain the real-time location information of the target vehicle; The scene information of the target vehicle's environment is determined based on the real-time location information.

6. The method according to any one of claims 1 to 3, characterized in that, The initial image acquired by the image acquisition device for the target vehicle, based on the environment in which the target vehicle is located, includes: When a preset gesture of a target user inside the target vehicle is detected, the wallpaper generation function is activated; When the wallpaper generation function is enabled, the initial image of the target vehicle is acquired by the image acquisition device for the environment in which the target vehicle is located.

7. The method according to any one of claims 1 to 3, characterized in that, The method further includes: The cabin wallpaper image is displayed on the in-vehicle infotainment system and / or ceiling screen of the target vehicle.

8. A device for generating cabin wallpaper, characterized in that, The device includes: An initial image acquisition module is used to acquire an initial image of the target vehicle captured by the image acquisition device of the target vehicle in relation to the environment in which the target vehicle is located; The information extraction module is used to determine the orientation information and attribute information of the target moving object based on the initial image; the attribute information includes the target moving object's own attributes and the target moving object's motion attributes; The prompt word generation module is used to generate target prompt words corresponding to the target image generation model based on the orientation information and the attribute information. The target image generation model is used to generate cockpit wallpapers of a preset style based on the prompt words and images. The cockpit wallpaper generation module is used to input the target prompt word and the initial image into the target image model and output the cockpit wallpaper image of the target vehicle. The information extraction module includes: The target bounding box information determination submodule is used to input the initial image into a motion object detection model for detecting moving objects and output the target bounding box information of the target moving object. The attribute information determination submodule is used to input the target bounding box information into the attribute recognition model for extracting the attribute information of the moving object, and output the attribute information of the target moving object; The depth estimation submodule is used to input the target bounding box information into a depth estimation model for estimating the depth information of a moving object, and output the depth information from the center point of the target bounding box of the moving object to the center point of the viewpoint of the image acquisition device. The orientation information determination submodule is used to determine the orientation information of the target moving object based on the target bounding box information and the depth information.

9. An electronic device, characterized in that, It includes a processor, a memory, and a computer program stored in the memory and capable of running on the processor, wherein the computer program, when executed by the processor, implements the method for generating cockpit wallpaper as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, A computer program is stored on the computer-readable storage medium, which, when executed by a processor, implements the method for generating cockpit wallpaper as described in any one of claims 1 to 7.

Images