Intelligent driving heavy fog scene generation method and system based on image processing and application thereof

Abstract

The application belongs to the technical field of image processing, and provides an intelligent driving heavy fog scene generation method and system based on image processing and application, comprising collecting a clear image of a real driving scene and generating a corresponding depth image; performing edge-aware smoothing processing on the depth image by using a weighted sparse gradient reconstruction algorithm to obtain a smoothed depth image, wherein the weighted sparse gradient reconstruction algorithm comprises constructing an energy minimization model and introducing a spatial adaptive weight matrix to sparsely constrain the gradient of the depth image; synthesizing a heavy fog scene image by using an atmospheric scattering model based on the smoothed depth image; and performing post-processing on the heavy fog scene image to complete the generation of the heavy fog scene. The application can realize efficient and high-precision image smoothing and be applied to heavy fog scene synthesis, thereby generating more real and reliable foggy day test images to provide efficient training and verification data for autonomous driving.

Description

Technical Field

[0001] This application relates to the field of image processing technology, and in particular to a method, system and application for generating fog scenes for intelligent driving based on image processing. Background Technology

[0002] The rapid development of intelligent driving technology relies heavily on the reliability of visual perception systems in complex environments. Fog, as the most common severe weather condition, causes atmospheric light scattering and absorption, leading to decreased image contrast, color distortion, and loss of detail. This severely weakens the perception capabilities of sensors such as cameras, making it particularly urgent to generate high-fidelity fog scene data in a controlled environment for system testing.

[0003] Currently, the testing and verification of autonomous driving systems mainly rely on two approaches: real-vehicle road testing and virtual simulation. Real-vehicle road testing is costly, time-consuming, and difficult to reproduce extreme foggy environments; while virtual simulation can generate various weather scenarios, balancing physical realism and computational efficiency in simulation remains a challenge for the industry. Real fog environments involve complex atmospheric optical processes such as Mie scattering, ambient light absorption, and transmittance changes, and traditional simplified fog effect simulation methods are insufficient to meet the high standards required for testing intelligent driving systems.

[0004] In terms of technical implementation, existing methods mainly rely on physically simulated rendering or generative adversarial networks (GANs), but still have many limitations. Physical simulation methods have high computational costs and poor adaptability to sensor characteristics; while image transformation methods based on GANs avoid complex physical modeling by learning an end-to-end mapping from clear images to foggy images, but their framework typically includes a generator and a discriminator. The generator often adopts an encoder-decoder structure and frequently preserves texture information through shortcut connections; the discriminator uses convolutional networks for real / fake classification. The loss function design usually combines GAN loss and pixel-level regression loss to improve generation accuracy. In recent years, optimized structures such as recurrent generative adversarial networks (RBANs) that utilize improved upsampling modules to reduce the checkerboard effect have also emerged.

[0005] Nevertheless, existing generation schemes still generally suffer from the following shortcomings:

[0006] 1. Regarding data dependency: GAN-type methods require a large number of paired clear-foggy images for supervised training, but the availability of precisely paired data in the real world is extremely limited. Unpaired training methods, on the other hand, suffer from poor model convergence stability.

[0007] 2. In terms of interpretability and controllability: As a "black box" model, GAN generates fog effects with unclear physical meaning, making it difficult to accurately control the fog's concentration, distribution, and other attributes. The control precision is far lower than that of physics-based models.

[0008] 3. In terms of detail preservation: GANs tend to lose subtle texture and edge information when processing high-resolution images, resulting in blurred or artifact-prone synthesized images. In particular, the quality of the generated images is not ideal when dealing with small targets at a distance, which affects the training effect of the perception algorithm.

[0009] 4. In terms of scene generation: Traditional simulation methods rely on the preset weather module and fixed map data in the scene generation platform, which cannot flexibly perform physical rendering on the data collected from actual roads under natural light. This results in low scene generation efficiency per person (usually less than 3 per day), serious scene homogenization, and insufficient realism of the generated fog effects due to the lack of depth information and dynamic interaction.

[0010] 5. At the image processing level: Traditional image fogging methods perform poorly in preserving object edges, resulting in blurred outlines of key objects in synthesized foggy images, directly impacting the target detection performance of intelligent driving systems. Furthermore, existing methods struggle to adaptively handle complex gradient changes in images, easily introducing unnatural lighting effects when simulating fog. In addition, intelligent driving systems have high real-time processing requirements, while existing algorithms are generally computationally inefficient, failing to meet the real-time demands of practical applications.

[0011] Therefore, a more efficient scene generation method is needed to address the shortcomings of existing image smoothing techniques in the synthesis of fog scenes for intelligent driving. Summary of the Invention

[0012] In view of the shortcomings of the prior art, the present invention provides a method, system and application for generating fog scenes for intelligent driving based on image processing. It can achieve efficient and high-precision image smoothing and be applied to fog scene synthesis, thereby generating more realistic and reliable fog test images and providing efficient training and verification data for autonomous driving.

[0013] To achieve the above and related objectives, the present invention adopts the following technical solution:

[0014] The first aspect of this invention provides a method for generating fog scenes for intelligent driving based on image processing, comprising the following steps:

[0015] Step S100: Acquire clear images of the real driving scene and generate corresponding depth images;

[0016] Step S200: The depth image is edge-aware smoothing is performed using a weighted sparse gradient reconstruction algorithm to obtain a smoothed depth image. The weighted sparse gradient reconstruction algorithm includes constructing an energy minimization model and introducing a spatial adaptive weight matrix to sparsely constrain the gradient of the depth image.

[0017] Step S300: Based on the smoothed depth image, synthesize a fog scene image using an atmospheric scattering model;

[0018] Step S400: Post-process the fog scene image to complete the fog scene generation.

[0019] Furthermore, in step S200, the energy minimization model includes:

[0020] (Formula 1),

[0021] In formula 1, Represents a depth image; This represents the smoothed depth image; Represents the image gradient; Represents the regularization parameter; Represents the gradient processing function; This represents the spatial adaptive weight matrix dynamically calculated based on the local gradient magnitude.

[0022] Furthermore, the spatially adaptive weight matrix includes:

[0023] (Formula 2),

[0024] In Formula 2, e represents the natural constant; σ represents the gradient magnitude; σ represents the scale parameter that controls the weight decay.

[0025] Furthermore, in the spatially adaptive weight matrix, the weight values ​​corresponding to the edge regions of the depth image are less than the weight values ​​corresponding to the smooth regions of the depth image.

[0026] Furthermore, in step S300, the atmospheric scattering model includes:

[0027] (Formula 3),

[0028] In formula 3, Image representing a foggy scene; A represents a sharp image; A represents global atmospheric light. This represents the transmittance, which is calculated based on the spatial distance data and scattering coefficient provided by the smoothed depth image.

[0029] Furthermore, the fog scene image includes scene geometry information, distance and depth information, and fog concentration and distribution information.

[0030] Furthermore, in step S400, post-processing includes image optimization of the foggy scene image to simulate the imaging characteristics of a real camera.

[0031] A second aspect of the present invention provides an image processing-based intelligent driving fog scene generation system, comprising:

[0032] The image preprocessing module is used to acquire clear images of real driving scenarios and generate corresponding depth images;

[0033] The smoothing module is used to perform edge-aware smoothing on the depth image using a weighted sparse gradient reconstruction algorithm to obtain a smoothed depth image. The weighted sparse gradient reconstruction algorithm includes constructing an energy minimization model and introducing a spatially adaptive weight matrix to sparsely constrain the gradient of the depth image.

[0034] The fog effect synthesis module is used to synthesize fog scene images based on smoothed depth images using an atmospheric scattering model.

[0035] The post-processing module is used to perform post-processing on fog scene images to complete the fog scene generation.

[0036] A third aspect of the present invention provides a computer-readable storage medium storing computer-readable instructions thereon, which, when executed by a computer's processor, cause the computer to perform the above-described image processing-based intelligent driving fog scene generation method.

[0037] A fourth aspect of the present invention provides a computer device, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the above-described intelligent driving fog scene generation method based on image processing.

[0038] The beneficial technical effects of this invention are as follows:

[0039] This invention extracts scenes from clear images of real-world driving scenarios and performs edge-aware smoothing on the processed depth images using a weighted sparse gradient reconstruction algorithm. This process suppresses noise while preserving the edge structure of the depth images. Furthermore, this invention utilizes an atmospheric scattering model for synthesis, which improves the physical realism, controllability, and generalization of foggy scene images. In addition, the spatially adaptive weight matrix of this invention can adaptively adjust weight values ​​according to the depth image content, making it suitable for various complex driving scenarios and generating more realistic fog effects.

[0040] This invention is applicable to autonomous driving simulation testing and can provide an effective fog synthesis scene for autonomous driving testing, thereby significantly reducing testing costs and improving rendering efficiency.

[0041] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and do not limit this application. Attached Figure Description

[0042] The accompanying drawings, incorporated in and forming part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application. It is obvious that the drawings described below are merely some embodiments of this application, and those skilled in the art can obtain other drawings based on these drawings without inventive effort. In the drawings:

[0043] Figure 1 This is a flowchart of the image processing-based intelligent driving fog scene generation method of this application;

[0044] Figure 2 This is a framework diagram of the image processing-based intelligent driving fog scene generation system of this application;

[0045] Figure 3 A schematic diagram of the structure of a computer system suitable for an embodiment of this application is shown. Detailed Implementation

[0046] Unless otherwise defined, all technical and / or scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. It should be understood that certain features of the invention (described in the context of separate embodiments for clarity) may also be provided in a single embodiment. Conversely, multiple features of the invention (described in the context of a single embodiment for brevity) may also be provided separately or in any suitable combination or, where appropriate, in any other described embodiment of the invention. Certain features described in the context of various embodiments will not be considered essential features of those embodiments unless the embodiment is inoperable without those elements. The invention is further illustrated below by specific examples; however, it should be noted that the specific process conditions and results described in the embodiments of the invention are merely illustrative and should not be construed as limiting the scope of protection of the invention. All equivalent changes or modifications made in accordance with the spirit and essence of the invention should be covered within the scope of protection of the invention.

[0047] Please see Figure 1 The flowchart of the image processing-based intelligent driving fog scene generation method of this application is described in detail below:

[0048] Step S100: Acquire clear images of the real driving scene and generate corresponding depth images.

[0049] Specifically, this application acquires images of driving scenes on real roads using sensors such as onboard cameras in intelligent driving vehicles. These images possess sufficient resolution and clarity, and contain multi-dimensional scene elements such as lane lines, vehicles, pedestrians, and traffic signs, enabling subsequent algorithms to accurately identify and process them. Then, this application can generate depth images based on various mature technologies such as stereo vision, monocular depth estimation, and LiDAR fusion. After generating the depth images, several key parameters are set for subsequent processing; these key parameters can include regularization parameters, scattering coefficients, and global atmospheric light.

[0050] Step S200: The depth image is edge-aware smoothing is performed using a weighted sparse gradient reconstruction algorithm to obtain a smoothed depth image. The weighted sparse gradient reconstruction algorithm includes constructing an energy minimization model and introducing a spatially adaptive weight matrix to sparsely constrain the gradient of the depth image.

[0051] Specifically, raw depth images typically contain significant noise, such as from sensor errors. Directly using them for fog scene synthesis can lead to distorted fog effects, such as mottled patterns. Furthermore, since the edges of depth images represent the contours of objects in a driving scene, excessive smoothing can blur these edges, causing fog to leak at object boundaries during subsequent fog effect synthesis. Therefore, this application performs edge-aware smoothing on depth images to avoid these problems. More specifically, the weighted sparse gradient reconstruction algorithm in this application achieves smoothing by constructing an energy-minimizing model. This model combines data fidelity terms and gradient sparsity terms, including:

[0052] (Formula 1),

[0053] In formula 1, Represents a depth image; This represents the smoothed depth image; Represents the image gradient; Represents the regularization parameter; Represents the gradient processing function; This represents the spatial adaptive weight matrix dynamically calculated based on the local gradient magnitude.

[0054] More specifically, in Formula 1, This represents the data fidelity factor, which measures the difference between the smoothed depth image and the original depth image. It ensures that the smoothed depth image does not deviate too far from the original depth image, thus preserving the basic structure of the real scene. This represents the gradient sparsity term, used to control the smoothness and edge preservation of depth images.

[0055] More specifically, the spatially adaptive weight matrix includes:

[0056] (Formula 2),

[0057] In Formula 2, e represents the natural constant; σ represents the gradient magnitude; σ represents the scale parameter controlling the weight decay. In the spatially adaptive weight matrix, the weight values ​​corresponding to edge regions of the depth image are small to preserve edge sharpness; the weight values ​​corresponding to smooth regions of the depth image are large to remove noise. Furthermore, the spatially adaptive weight matrix of this application can adaptively adjust the weight values ​​according to the depth image content, making it suitable for various complex scenarios, such as urban roads and highways.

[0058] Step S300: Based on the smoothed depth image, synthesize a fog scene image using an atmospheric scattering model.

[0059] Specifically, the atmospheric scattering model of this application includes:

[0060] (Formula 3),

[0061] In formula 3, Image representing a foggy scene; A represents a sharp image; A represents global atmospheric light. This represents the transmittance, which is calculated based on the spatial distance data and scattering coefficient provided by the smoothed depth image.

[0062] More specifically, the formula for calculating transmittance is as follows:

[0063] (Formula 4),

[0064] In Formula 4, d(x) represents the smoothed depth image; β represents the scattering coefficient; and e represents the natural constant.

[0065] More specifically, this application uses an atmospheric scattering model to describe the physical process of light propagation in foggy conditions. During propagation, light is scattered and absorbed by particles in the fog. The light that finally reaches the camera consists of two parts: the first part is directly transmitted light, i.e., the clear image. The pixel brightness is calculated as follows: the first part is the light intensity directly reaching the camera from the object, after attenuation by fog; the second part is atmospheric scattered light, which is the background light intensity scattered by fog after a portion of the global atmospheric light A is scattered by the fog. Finally, the direct transmitted light and the atmospheric scattered light are added together to obtain the final brightness G(x) of the pixel in foggy weather. The smoothed depth image provides spatial distance information, simulating the fog concentration distribution. A smaller d(x) results in a larger t(x), making nearby objects relatively clear; a larger d(x) results in a smaller t(x), making distant objects relatively blurry, thus simulating the rule that visibility decreases with distance in real foggy weather. The scattering coefficient is used to control fog concentration. At the same depth, a larger scattering coefficient results in a faster decrease in transmittance, an increased proportion of atmospheric scattered light, a more blurred overall image, and denser fog; a smaller scattering coefficient results in a slower decrease in transmittance, an increased proportion of direct transmitted light, a clearer overall image, and lighter fog. Therefore, the fog scene image synthesized in this application is more realistic.

[0066] More specifically, the fog scene images in this application include scene geometry information, distance and depth information, and fog concentration and distribution information.

[0067] Step S400: Post-process the fog scene image to complete the fog scene generation.

[0068] Specifically, post-processing includes image optimization of foggy scene images to simulate the imaging characteristics of a real camera. Image optimization in this application includes, but is not limited to, contrast adjustment, to make foggy scene images visually closer to photographs taken on real foggy days.

[0069] More specifically, this application can integrate post-processed fog scene images into an intelligent driving test platform for the evaluation of object detection and scene understanding algorithms.

[0070] More specifically, the method of this application is not limited to the synthesis of fog scenes in intelligent driving, but can also be extended to other fields, such as virtual reality (VR) and augmented reality (AR), to generate realistic weather effects and improve user experience; autonomous driving simulation can be extended to the synthesis of other weather conditions, such as rain and snow, to provide a comprehensive testing environment for autonomous driving.

[0071] Please see Figure 2 Here is a framework diagram of the image processing-based intelligent driving fog scene generation system 200 of this application, including:

[0072] The image preprocessing module 210 is used to acquire clear images of real driving scenarios and generate corresponding depth images;

[0073] The smoothing module 220 is used to perform edge-aware smoothing on the depth image using a weighted sparse gradient reconstruction algorithm to obtain a smoothed depth image. The weighted sparse gradient reconstruction algorithm includes constructing an energy minimization model and introducing a spatially adaptive weight matrix to sparsely constrain the gradient of the depth image.

[0074] Fog effect synthesis module 230 is used to synthesize fog scene images based on smoothed depth images using an atmospheric scattering model;

[0075] The post-processing module 240 is used to perform post-processing on the fog scene image to complete the fog scene generation.

[0076] It should be noted that the intelligent driving fog scene generation system and the intelligent driving fog scene generation method provided in the above embodiments belong to the same concept. The specific methods of operation of each module and unit have been described in detail in the method embodiments and will not be repeated here. In practical applications, the intelligent driving fog scene generation system provided in the above embodiments can be assigned to different functional modules as needed, that is, the internal structure of the system can be divided into different functional modules to complete all or part of the functions described above. This is not a limitation here.

[0077] Embodiments of this application also provide a computer device, including: one or more processors; and a storage device for storing one or more programs, which, when executed by the one or more processors, cause the computer device to implement the intelligent driving fog scene generation method provided in the above embodiments.

[0078] Figure 3 A schematic diagram of the structure of a computer system suitable for an embodiment of this application is shown. It should be noted that... Figure 3 The computer system 300 of the electronic device shown is merely an example and should not impose any limitation on the functionality and scope of use of the embodiments of this application.

[0079] like Figure 3As shown, the computer system 300 includes a central processing unit (CPU) 301, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 302 or a program loaded from a storage section 308 into a random access memory (RAM) 303, such as performing the methods described in the above embodiments. Various programs and data required for system operation are also stored in the RAM 303. The CPU 301, ROM 302, and RAM 303 are interconnected via a bus 304. An input / output (I / O) interface 305 is also connected to the bus 304. The following components are connected to the I / O interface 305: an input section 306 including a keyboard, mouse, etc.; an output section 307 including a cathode ray tube (CRT), liquid crystal display (LCD), etc., and speakers, etc.; a storage section 308 including a hard disk, etc.; and a communication section 309 including a network interface card such as a LAN (local area network) card, modem, etc. The communication section 309 performs communication processing via a network such as the Internet. A driver 310 is also connected to the I / O interface 305 as needed. Removable media 311, such as disks, optical discs, magneto-optical discs, semiconductor memories, etc., are installed on drive 310 as needed so that computer programs read from them can be installed into storage section 308 as needed.

[0080] Specifically, according to embodiments of this application, the processes described above with reference to the flowcharts can be implemented as computer tool programs. For example, embodiments of this application include a computer program product comprising a computer program carried on a computer-readable medium, the computer program including a computer program for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via communication section 309, and / or installed from removable medium 311. When the computer program is executed by central processing unit (CPU) 301, it performs various functions defined in the system of this application.

[0081] It should be noted that the computer-readable medium shown in the embodiments of this application can be a computer-readable signal medium or a computer-readable storage medium, or any combination thereof. A computer-readable storage medium can be, for example, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of a computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer disk, a hard disk, a random access memory, a read-only memory, an erasable programmable read-only memory, flash memory, an optical fiber, a portable compact disk read-only memory, an optical storage device, a magnetic storage device, or any suitable combination thereof. In this application, a computer-readable signal medium can include a data signal propagated in baseband or as part of a carrier wave, carrying a computer-readable computer program. Such propagated data signals can take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A computer-readable signal medium can also be any computer-readable medium other than a computer-readable storage medium, which can send, propagate, or transmit a program for use by or in conjunction with an instruction execution system, apparatus, or device. Computer programs contained on computer-readable media can be transmitted using any suitable medium, including but not limited to wireless, wired, etc., or any suitable combination thereof.

[0082] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this application. Each block in a flowchart or block diagram may represent a module, segment, or portion of code, which contains one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in a block diagram or flowchart, and combinations of blocks in a block diagram or flowchart, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0083] The units described in the embodiments of this application can be implemented by tools or by hardware, and the described units can also be located in a processor. The names of these units do not necessarily limit the unit itself.

[0084] Another aspect of this application provides a computer-readable storage medium storing a computer program thereon, which, when executed by a computer's processor, causes the computer to perform the intelligent driving fog scene generation method described above. This computer-readable storage medium may be included in the computer device described in the above embodiments, or it may exist independently and not assembled into the computer device.

[0085] Another aspect of this application provides a computer program product or computer program including computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the computer device to perform the intelligent driving fog scene generation method provided in the various embodiments described above.

[0086] The above embodiments are merely illustrative of the principles and effects of the present invention and are not intended to limit the invention. Any person skilled in the art can modify or alter the above embodiments without departing from the spirit and scope of the present invention. Therefore, all equivalent modifications or alterations made by those skilled in the art without departing from the spirit and technical concept disclosed in the present invention should still be covered by the claims of the present invention.

Claims

1. A method for generating fog scenes for intelligent driving based on image processing, characterized in that, Includes the following steps: Step S100: Acquire clear images of the real driving scene and generate corresponding depth images; Step S200: The depth image is processed by edge-aware smoothing using a weighted sparse gradient reconstruction algorithm to obtain a smoothed depth image. The weighted sparse gradient reconstruction algorithm includes constructing an energy minimization model and introducing a spatially adaptive weight matrix to sparsely constrain the gradient of the depth image. Step S300: Based on the smoothed depth image, synthesize a fog scene image using an atmospheric scattering model; Step S400: Post-process the fog scene image to complete the fog scene generation.

2. The generation method according to claim 1, characterized in that, In step S200, the energy minimization model includes: (Official 1), In formula 1, This represents the depth image; This represents the smoothed depth image; Represents the image gradient; Represents the regularization parameter; Represents the gradient processing function; This represents the spatial adaptive weight matrix dynamically calculated based on the local gradient magnitude.

3. The generation method according to claim 2, characterized in that, The spatial adaptive weight matrix includes: (Official 2), In Formula 2, e represents the natural constant; σ represents the gradient magnitude; σ represents the scale parameter that controls the weight decay.

4. The generation method according to claim 3, characterized in that, In the spatial adaptive weight matrix, the weight values ​​corresponding to the edge regions of the depth image are less than the weight values ​​corresponding to the smooth regions of the depth image.

5. The generation method according to claim 1, characterized in that, In step S300, the atmospheric scattering model includes: (Official 3), In formula 3, This represents the image of the foggy scene; The image represents the clear image; A represents global atmospheric light; The transmittance is represented by the spatial distance data and scattering coefficient provided by the smoothed depth image.

6. The generation method according to claim 5, characterized in that, The fog scene image includes scene geometry information, distance and depth information, fog concentration and distribution information.

7. The generation method according to claim 1, characterized in that, In step S400, the post-processing includes image optimization of the foggy scene image to simulate the imaging characteristics of a real camera.

8. An intelligent driving fog scene generation system based on image processing, characterized in that, include: The image preprocessing module is used to acquire clear images of real driving scenarios and generate corresponding depth images; The smoothing module is used to perform edge-aware smoothing on the depth image using a weighted sparse gradient reconstruction algorithm to obtain a smoothed depth image. The weighted sparse gradient reconstruction algorithm includes constructing an energy minimization model and introducing a spatially adaptive weight matrix to sparsely constrain the gradient of the depth image. The fog effect synthesis module is used to synthesize a fog scene image based on the smoothed depth image using an atmospheric scattering model. The post-processing module is used to perform post-processing on the fog scene image to complete the fog scene generation.

9. A computer-readable storage medium, characterized in that, It stores computer-readable instructions, which, when executed by the computer's processor, cause the computer to perform the image processing-based intelligent driving fog scene generation method according to any one of claims 1 to 7.

10. A computer device, comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, when the processor executes the computer program, it implements the steps of the image processing-based intelligent driving fog scene generation method as described in any one of claims 1 to 7.

Images