System and method for quickly generating panoramic dynamic graph on line based on 3D model

Abstract

The invention discloses a system and method for quickly generating a panoramic dynamic graph on line based on a 3D model, and the method comprises the following steps: S1, obtaining 3D model data, and generating a 3D model; s2, simplifying the 3D model, and obtaining different view angle images by using the 3D model; s3, synthesizing the images of different visual angles into a panoramic image through an improved gradual transition algorithm; s4, updating the illumination effect of the panoramic image according to the illumination intensity; s5, generating a dynamic effect for the panoramic image by adopting a particle system technology; s6, performing rendering processing through a cloud cluster, and dynamically adjusting an image compression ratio by adopting an improved adaptive compression technology; and S7, the rendering angle and the dynamic effect of the panoramic image are adjusted in real time. According to the method, an efficient and scientific optimization scheme can be provided in 3D image generation, and remarkable technical values and economic benefits are brought to practical application.

Description

Technical Field

[0001] The present invention relates to the technical field of 3D modeling and image rendering, and in particular to a system and method for quickly generating a panoramic dynamic image online based on a 3D model. Background Art

[0002] With the development of computer graphics and virtual reality technology, generating panoramic dynamic images based on 3D models has become a key technology in many application fields and is widely used in multiple industries such as games, film production, virtual reality, and augmented reality. Panoramic dynamic images not only need to display the details of the three-dimensional model from multiple perspectives, but also require high-quality images and smooth user interaction experience during the rendering process. In order to achieve these goals, existing technologies rely on a series of complex rendering and optimization algorithms, usually combined with technologies such as 3D modeling, dynamic lighting, particle systems, and image compression.

[0003] Most existing panoramic image generation methods focus on how to generate static panoramic images or process relatively simple dynamic scenes. Traditional image generation methods usually use static 3D models to generate panoramic images and render them through fixed rendering rules. This method often has a large bottleneck when processing dynamic perspectives or complex image effects. For example, when faced with a large number of perspective changes, existing image rendering technologies often find it difficult to balance rendering efficiency and image quality, resulting in increased latency in the image generation process, especially in resource-constrained devices or network environments. In addition, although image compression technology can reduce the amount of transmitted data to a certain extent, it often affects image quality, especially when network bandwidth is limited or image detail requirements are high, the efficiency and adaptability of the compression algorithm are still insufficient.

[0004] Among the existing panoramic dynamic image generation methods, although some technologies have been able to achieve dynamic perspective adjustment and real-time image rendering, these methods still face great challenges in balancing latency, image quality and transmission efficiency when dealing with user interactions. For example, many existing systems cannot adapt to different network conditions and device performance, resulting in limited image quality. Especially in dynamic interactions, the details and smoothness of the image often cannot meet the user's experience requirements. In addition, traditional rendering and compression technologies often cannot adaptively adjust the compression ratio or image rendering strategy when the network bandwidth is low or the device performance is weak, resulting in delays and screen freezes during use.

[0005] At present, image generation methods based on cloud-based distributed rendering architecture are also gradually being applied to this field. Cloud-based rendering achieves efficient parallel computing by distributing computing tasks to multiple computing nodes, which can effectively improve the rendering speed. However, existing cloud-based rendering technologies often have data transmission bottlenecks when processing highly complex images. Real-time transmission and rendering of images require a lot of bandwidth and computing resources, which is still a huge challenge for environments with poor device performance or unstable networks. Especially for large-scale panoramic dynamic image generation, how to ensure rendering quality while avoiding image distortion or delays caused by bandwidth limitations is still a problem that needs to be solved urgently.

[0006] In view of the defects of the above-mentioned prior art, the present invention proposes a system and method for quickly generating panoramic dynamic images online based on 3D models. The method aims to achieve high-quality generation and efficient transmission of panoramic dynamic images by introducing dynamic lighting calculation, particle system, distributed rendering architecture and adaptive image compression technology. Compared with traditional technologies, the present invention not only optimizes the dynamic adjustment of viewing angle and image detail enhancement during the rendering process, but also automatically adjusts the image compression ratio and rendering strategy according to device performance and network conditions, effectively solving the problems of image generation delay, low transmission efficiency and quality loss in the prior art. In addition, based on the cloud-based distributed rendering architecture, the present invention can achieve efficient parallel computing through dynamic scheduling and load balancing in a resource-constrained environment, ensuring that users get a smooth and delay-free panoramic dynamic image experience during the interaction process.

[0007] Therefore, the present invention not only solves the performance bottleneck and user experience problems faced by the prior art, but also provides a more flexible and efficient solution for future panoramic dynamic image generation and rendering technology. Summary of the invention

[0008] One purpose of the present invention is to propose a system and method for quickly generating panoramic dynamic images online based on 3D models. The present invention can provide an efficient and scientific optimization solution in the generation of panoramic dynamic images, bringing significant technical value and economic benefits to practical applications.

[0009] According to the present invention, a method for quickly generating a panoramic dynamic image online based on a 3D model comprises the following steps: S1, obtaining 3D model data and generating a 3D model through a 3D modeling tool; S2, simplifying the 3D model, and using the 3D model to obtain images from different viewing angles; S3, synthesizing the images of different viewing angles into a panoramic image by using an improved gradual transition algorithm, wherein the improved gradual transition algorithm performs a smooth transition of color and texture on the spliced ​​area; S4, performing real-time calculation of illumination, shadow, and reflection in the panoramic image, and updating the illumination effect of the panoramic image according to the illumination intensity; S5. Use particle system technology to generate dynamic effects for panoramic images; S6. Use the cloud distributed rendering architecture to distribute image generation tasks to multiple computing nodes, perform rendering processing through the cloud cluster, and use improved adaptive compression technology to dynamically adjust the image compression ratio in real time according to the user's network conditions and device performance; S7. According to the user's operation instruction, the rendering angle and dynamic effect of the panoramic image are adjusted in real time.

[0010] Optionally, the S2 specifically includes: S21. Use the QEM algorithm to simplify the mesh of the 3D model. The QEM algorithm calculates the geometric error of each triangular face. The geometric error is represented by an error matrix Q. The error matrix Q is the distance error between the vertex coordinates of each triangular face and the geometric center point of the adjacent face of the vertex coordinates: ; in, is the coordinate of each vertex on the triangle patch, is the geometric center coordinate of the triangle patch, and n is the number of vertices; The error matrix Q of each triangular face is applied to the simplification process. The triangular face with the smallest error is selected for simplification according to the error size. The geometric position of the simplified new triangular face is the weighted average of its error matrix Q. Iteratively update the simplified new triangle patches, and continue to select triangle patches with the smallest error for merging until the preset number of triangle patches M is reached; S22, simplifying the texture map of the 3D model, and obtaining a texture block size T based on the number of triangle patches M; S23. Evaluate the error between the simplified model and the original model using the error metric to obtain the error threshold , if the error is less than the preset threshold, the simplified model is judged to be qualified.

[0011] Optionally, the S3 specifically includes: S31, using an improved gradual transition algorithm to smoothly transition multiple perspective images, wherein the improved gradual transition algorithm uses a gradual factor G(x, y) for smoothing: ; Among them, d(x,y) represents the distance between the pixel point (x,y) in the stitching area and the stitching edge. is the maximum distance, C(x,y) represents the color value of the pixel (x,y), is the average color of the area, , is the regulating factor, is the natural exponential function; S32. Obtain a smoothed viewing angle image according to the gradient factor G(x,y): ; in, is the smoothed pixel value, is the pixel value of the previous image. is the pixel value of the image after stitching; S33, synthesizing the multiple perspective images into a panoramic image.

[0012] Optionally, the S4 specifically includes: S41, real-time calculation of the illumination intensity I(x,y) in the panoramic image: ; in, is ambient light, is diffuse reflected light, is the specular reflected light, To refract light, For scattered light; The diffusely reflected light The calculation is as follows: ; in, is the influence coefficient of the incident angle of light on diffuse reflection, is the dot product of the surface normal and the light source direction, is the intensity of the light source; The specular reflected light The calculation is as follows: ; in, is the mirror reflection coefficient, H is the half-distance vector, is the roughness index, V is the observer direction vector, which represents the unit vector pointing from the observer to the surface being rendered, is the dot product of the surface normal and the light source direction, is the dot product of the surface normal and the half-range vector, is the intensity of the light source; The refracted light The calculation formula is: ; in, is the Fresnel-Schlick approximation coefficient, is the refractive index of the medium, is the dot product of the surface normal and the light source direction, is the intensity of the light source; The scattered light The calculation formula is: ; in, and are the Rayleigh and Mie scattering coefficients, is the dot product of the surface normal and the light source direction, is the adjustment coefficient, is the intensity of the light source; S42: Update the lighting effect in the panoramic image according to the light intensity I(x, y) calculated in real time.

[0013] Optionally, the S5 specifically includes: S51. Apply particle system technology to generate dynamic effects in the panoramic image. The particle system is composed of a number of particles. The state of each particle is composed of position P(x, y, z), velocity V(x, y, z), acceleration A(x, y, z) and life cycle Definition: When generating particles, random distribution and gradient weighting methods are used. The initial position of the particles Randomization is performed based on the spatial distribution within the region, and when the particle position is updated, the Gaussian distribution is used to model the particle movement; S52. The update formula of the particle system is: ; ; in, is the new position of the particle, is the initial position of the particle, is the new velocity of the particle, is the initial velocity of the particle, is the time step; S53. According to the generated dynamic particle effect, the movement trajectory of the particles is adjusted in real time, and the generated dynamic effect is superimposed on the panoramic image.

[0014] Optionally, the S6 specifically includes: S61, using cloud-based distributed rendering architecture, distributes image rendering tasks to multiple computing nodes, and the rendering nodes dynamically allocate tasks based on the load balancing algorithm: ; in, is the computing task of the i-th rendering node, is the load of the i-th rendering node, is the computing power of the i-th rendering node, is the bandwidth, N is the total number of nodes, is the load of the ith node, The computing power of the ith node, is the bandwidth of the i-th node; S62. Obtain the rendering priority of each image block: ; in, is the priority of the image block, is the rendering time of the i-th image block, is the spatial complexity of the i-th image block, , is the regulating factor; S63. Adopt a rendering task transmission strategy based on multi-level network bandwidth control to dynamically adjust the bandwidth of data transmission: ; in, Bandwidth allocation for transmission tasks, is the maximum bandwidth, is the priority of the transmission task, M is the number of tasks; S64, providing real-time feedback and adjustment on the calculation results of each rendering node, calculating the workload of the rendering node through the feedback mechanism and adjusting the task allocation between the nodes, the feedback adjustment formula is: ; in, is the adjusted rendering task volume of the i-th rendering node, is the computing task of the i-th rendering node, is the global average load, is the load of the i-th rendering node, is the regulating factor; S65, using improved adaptive compression technology, dynamically adjust the image compression ratio in real time according to the user's network conditions and device performance to adapt to different bandwidths and device capabilities: ; in, is the compression ratio, is the transmission task bandwidth, is the priority of the image block, is the rendering task volume, is the node load, is the number of image blocks, is the image complexity, , is the adjustment coefficient, is the adaptive adjustment function, is the natural exponential function; S66, according to the compression ratio Compressing panoramic images; S67, combined with the quality evaluation of compressed images, optimize the adaptive compression algorithm, and the quality of compressed images for: ; in, is the entropy value of the image, For data transmission delay, is the maximum allowed delay, is the delay influence coefficient, is the complexity influence coefficient, is the image complexity.

[0015] According to an embodiment of the present invention, a system for quickly generating a panoramic dynamic image online based on a 3D model includes the following modules: 3D model acquisition and processing module, used to collect 3D model data and simplify and optimize it; A panoramic image synthesis module is used to synthesize a panoramic image from multiple perspective images through cubic projection, and to calculate the overlapping part of the stitching area using image stitching technology; Dynamic lighting and effect generation module, used to calculate lighting, shadows, reflections and dynamic effects in panoramic images to generate natural dynamic effects; The rendering and distributed computing module distributes rendering tasks to multiple computing nodes through the cloud distributed rendering architecture, dynamically optimizes task allocation based on node load, bandwidth, and computing power, and performs parallel processing and optimization of rendering results. Adaptive compression and transmission optimization module, which is used to adjust the image compression ratio according to the user's network bandwidth and device performance, using adaptive compression technology to balance image quality and transmission efficiency, while reducing transmission delay through network optimization strategies; The real-time feedback and optimization module is used to provide feedback and adjust image details and resolution based on real-time rendering results.

[0016] The beneficial effects of the present invention are: First of all, the present invention solves the performance bottlenecks and user experience problems faced in the process of panoramic dynamic image generation and rendering in the prior art by innovatively introducing multiple technologies. In terms of image generation, dynamic lighting calculation and particle system based on 3D models are adopted, which makes the panoramic dynamic image more realistic and delicate in terms of lighting, shadow, reflection and dynamic effects, and improves the visual effect of the image. Compared with traditional static images or simple dynamic images, the introduction of dynamic lighting enables the image to be adaptively adjusted according to changes in light source and scene at different viewing angles, thereby enhancing the user's sense of immersion.

[0017] Secondly, the present invention optimizes the computing process of image rendering through a cloud-based distributed rendering architecture, distributes rendering tasks to multiple computing nodes, and improves rendering efficiency. The system can dynamically adjust the rendering load according to the network environment and device performance, avoiding rendering delays caused by insufficient computing power of a single node, and ensuring the high efficiency of large-scale panoramic dynamic image generation. This innovation effectively reduces the delays caused by computing bottlenecks in traditional rendering methods, and is particularly suitable for devices and network environments with limited resources.

[0018] In addition, the present invention introduces adaptive compression technology, which can automatically adjust the compression ratio and rendering details of the image according to the real-time network bandwidth and device performance. This allows the image to maintain a high quality and reduce transmission delay even in a low-bandwidth or low-performance device environment. The adaptive compression algorithm can optimize the transmission efficiency while ensuring image quality, avoiding quality loss and stuttering in traditional compression methods.

[0019] Finally, the perspective dynamic adjustment technology and real-time feedback optimization module of the present invention ensure efficient response during user interaction. When the user changes the perspective or operates, the system can calculate the new perspective in real time and optimize image rendering to ensure that the image quality and smoothness are not affected. Through these optimization measures, the present invention can provide a smoother and more natural user experience when processing complex dynamic images. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings: Figure 1 This is an overall flow chart of a method for quickly generating a panoramic dynamic image online based on a 3D model proposed by the present invention; Figure 2 This is a structural schematic diagram of a system for quickly generating panoramic dynamic images online based on a 3D model proposed by the present invention. DETAILED DESCRIPTION

[0021] The present invention will now be described in further detail with reference to the accompanying drawings. These drawings are simplified schematic diagrams, which only illustrate the basic structure of the present invention in a schematic manner, and therefore only show the components related to the present invention.

[0022] refer to Figure 1 A method for quickly generating a panoramic dynamic image online based on a 3D model comprises the following steps: S1, obtaining 3D model data and generating a 3D model through a 3D modeling tool; S2, simplifying the 3D model, and using the 3D model to obtain images from different viewing angles; S3, synthesizing the images of different viewing angles into a panoramic image by using an improved gradual transition algorithm, wherein the improved gradual transition algorithm performs a smooth transition of color and texture on the spliced ​​area; S4, performing real-time calculation of illumination, shadow, and reflection in the panoramic image, and updating the illumination effect of the panoramic image according to the illumination intensity; S5. Use particle system technology to generate dynamic effects for panoramic images; S6. Use the cloud distributed rendering architecture to distribute image generation tasks to multiple computing nodes, perform rendering processing through the cloud cluster, and use improved adaptive compression technology to dynamically adjust the image compression ratio in real time according to the user's network conditions and device performance; S7. According to the user's operation instruction, the rendering angle and dynamic effect of the panoramic image are adjusted in real time.

[0023] In this implementation, S2 specifically includes: S21. Use the QEM algorithm to simplify the mesh of the 3D model. The QEM algorithm calculates the geometric error of each triangular face. The geometric error is represented by an error matrix Q. The error matrix Q is the distance error between the vertex coordinates of each triangular face and the geometric center point of the adjacent face of the vertex coordinates: ; in, is the coordinate of each vertex on the triangle patch, is the geometric center coordinate of the triangle patch, and n is the number of vertices; The error matrix Q of each triangular face is applied to the simplification process. The triangular face with the smallest error is selected for simplification according to the error size. The geometric position of the simplified new triangular face is the weighted average of its error matrix Q. Iteratively update the simplified new triangle patches, and continue to select triangle patches with the smallest error for merging until the preset number of triangle patches M is reached; S22, simplifying the texture map of the 3D model, and obtaining a texture block size T based on the number of triangle patches M; S23. Evaluate the error between the simplified model and the original model using the error metric to obtain the error threshold , if the error is less than the preset threshold, the simplified model is judged to be qualified.

[0024] In this implementation, S3 specifically includes: S31, using an improved gradual transition algorithm to smoothly transition multiple perspective images, wherein the improved gradual transition algorithm uses a gradual factor G(x, y) for smoothing: ; Among them, d(x,y) represents the distance between the pixel point (x,y) in the stitching area and the stitching edge. is the maximum distance, C(x,y) represents the color value of the pixel (x,y), is the average color of the area, , is the regulating factor, is the natural exponential function; S32. Obtain a smoothed viewing angle image according to the gradient factor G(x,y): ; in, is the smoothed pixel value, is the pixel value of the previous image. is the pixel value of the image after stitching; S33, synthesizing the multiple perspective images into a panoramic image.

[0025] In this implementation manner, the S4 specifically includes: S41, real-time calculation of the illumination intensity I(x,y) in the panoramic image: ; in, is ambient light, is diffuse reflected light, is the specular reflected light, To refract light, For scattered light; The diffusely reflected light The calculation is as follows: ; in, is the influence coefficient of the incident angle of light on diffuse reflection, is the dot product of the surface normal and the light source direction, is the intensity of the light source; The specular reflected light The calculation is as follows: ; in, is the mirror reflection coefficient, H is the half-distance vector, is the roughness index, V is the observer direction vector, which represents the unit vector pointing from the observer to the surface being rendered, is the dot product of the surface normal and the light source direction, is the dot product of the surface normal and the half-range vector, is the intensity of the light source; The refracted light The calculation formula is: ; in, is the Fresnel-Schlick approximation coefficient, is the refractive index of the medium, is the dot product of the surface normal and the light source direction, is the intensity of the light source; The scattered light The calculation formula is: ; in, and are the Rayleigh and Mie scattering coefficients, is the dot product of the surface normal and the light source direction, is the adjustment coefficient, is the intensity of the light source; S42: Update the lighting effect in the panoramic image according to the light intensity I(x, y) calculated in real time.

[0026] In this implementation manner, S5 specifically includes: S51. Apply particle system technology to generate dynamic effects in the panoramic image. The particle system is composed of a number of particles. The state of each particle is composed of position P(x, y, z), velocity V(x, y, z), acceleration A(x, y, z) and life cycle Definition: When generating particles, random distribution and gradient weighting methods are used. The initial position of the particles Randomization is performed based on the spatial distribution within the region, and when the particle position is updated, the Gaussian distribution is used to model the particle movement; S52. The update formula of the particle system is: ; ; in, is the new position of the particle, is the initial position of the particle, is the new velocity of the particle, is the initial velocity of the particle, is the time step; S53. According to the generated dynamic particle effect, the movement trajectory of the particles is adjusted in real time, and the generated dynamic effect is superimposed on the panoramic image.

[0027] In this implementation manner, S6 specifically includes: S61, using cloud-based distributed rendering architecture, distributes image rendering tasks to multiple computing nodes, and the rendering nodes dynamically allocate tasks based on the load balancing algorithm: ; in, is the computing task of the i-th rendering node, is the load of the i-th rendering node, is the computing power of the i-th rendering node, is the bandwidth, N is the total number of nodes, is the load of the ith node, The computing power of the ith node, is the bandwidth of the i-th node; S62. Obtain the rendering priority of each image block: ; in, is the priority of the image block, is the rendering time of the i-th image block, is the spatial complexity of the i-th image block, , is the regulating factor; S63. Adopt a rendering task transmission strategy based on multi-level network bandwidth control to dynamically adjust the bandwidth of data transmission: ; in, Bandwidth allocation for transmission tasks, is the maximum bandwidth, is the priority of the transmission task, M is the number of tasks; S64, providing real-time feedback and adjustment on the calculation results of each rendering node, calculating the workload of the rendering node through the feedback mechanism and adjusting the task allocation between the nodes, the feedback adjustment formula is: ; in, is the adjusted rendering task volume of the i-th rendering node, is the computing task of the i-th rendering node, is the global average load, is the load of the i-th rendering node, is the regulating factor; S65, using improved adaptive compression technology, dynamically adjust the image compression ratio in real time according to the user's network conditions and device performance to adapt to different bandwidths and device capabilities: ; in, is the compression ratio, is the transmission task bandwidth, is the priority of the image block, is the rendering task volume, is the node load, is the number of image blocks, is the image complexity, , is the adjustment coefficient, is the adaptive adjustment function, is the natural exponential function; S66, according to the compression ratio Compressing panoramic images; S67, combined with the quality evaluation of compressed images, optimize the adaptive compression algorithm, and the quality of compressed images for: ; in, is the entropy value of the image, For data transmission delay, is the maximum allowed delay, is the delay influence coefficient, is the complexity influence coefficient, is the image complexity.

[0028] refer to Figure 2 , a system for quickly generating panoramic dynamic images online based on 3D models, including the following modules: 3D model acquisition and processing module, used to collect 3D model data and simplify and optimize it; A panoramic image synthesis module is used to synthesize a panoramic image from multiple perspective images through cubic projection, and to calculate the overlapping part of the stitching area using image stitching technology; Dynamic lighting and effect generation module, used to calculate lighting, shadows, reflections and dynamic effects in panoramic images to generate natural dynamic effects; The rendering and distributed computing module distributes rendering tasks to multiple computing nodes through the cloud distributed rendering architecture, dynamically optimizes task allocation based on node load, bandwidth, and computing power, and performs parallel processing and optimization of rendering results. Adaptive compression and transmission optimization module, which is used to adjust the image compression ratio according to the user's network bandwidth and device performance, using adaptive compression technology to balance image quality and transmission efficiency, while reducing transmission delay through network optimization strategies; The real-time feedback and optimization module is used to provide feedback and adjust image details and resolution based on real-time rendering results.

[0029] Embodiment 1: Embodiment In an online education platform, with the rapid development of online education, virtual laboratories, as an innovative teaching tool, have gradually become an important part of many education platforms. Through virtual laboratories, students can perform various experimental operations without actual equipment, master practical operation skills and conduct interactive learning. With the advancement of technology, the interactivity, realism and diversified experimental scene requirements of virtual laboratories are also increasing. Traditional virtual laboratory technology often faces problems such as image delay, slow loading and low image quality, which affect students' immersion and learning efficiency.

[0030] Suppose an online education platform offers a physics course that includes various classic mechanics experiments, such as the inclined plane slider experiment and the spring oscillator experiment. These experiments require precise physical modeling and interactive control. Students can simulate and operate these experimental equipment in a virtual environment through a virtual laboratory. In the virtual laboratory, students can adjust experimental parameters, change experimental conditions and view experimental results by mouse clicks, keyboard operations or virtual reality devices.

[0031] In the application of this virtual laboratory, all experimental equipment (such as inclined planes, sliders, springs, etc.) are constructed in the form of 3D models and rendered in the background through the system of the present invention for quickly generating panoramic dynamic images online based on 3D models. When students conduct experiments, each operation will affect the experimental results in real time and be reflected in the virtual laboratory.

[0032] Through the cloud-based distributed rendering architecture of the present invention, all rendering tasks are assigned to multiple computing nodes, and the load balancing algorithm is used to optimize the rendering speed, reduce image delays, and ensure that each student's operation can be quickly reflected in the experimental scene. At the same time, in order to adapt to different network conditions and the performance of student equipment, the system adopts adaptive compression technology to reduce data transmission delays and bandwidth pressure while ensuring the quality of experimental images, ensuring smooth loading and real-time updating of images. During the experiment, students can adjust the parameters of the experimental equipment through simple interactions (such as dragging the mouse and clicking buttons). Whenever the viewing angle changes or the experimental parameters change, the system will quickly update the image through the viewing angle dynamic adjustment module and the rendering module according to the student's operation to ensure instant feedback of the operation.

[0033] In order to verify the beneficial effects of the present invention, the implementers conducted experimental tests in a real online education platform. The experimental scenario was a student conducting an inclined slider experiment in a virtual laboratory. In this experiment, the student could drag the slider with the mouse and change the angle of the inclined plane to observe the movement of the slider. The implementers tested the image rendering speed, loading delay and image quality of the experiment under different network bandwidth conditions and device configurations.

[0034] By comparing with the traditional method, the specific data are as follows Table 1: Table 1 Comparison of image rendering speed and loading delay results between the traditional method and the method of the present invention ; In the entire embodiment, the implementer not only solves the problems of image rendering delay, slow loading and low image quality in the virtual laboratory through the method of the present invention, but also improves the interaction efficiency of the laboratory system, and realizes the efficient operation of the virtual experiment and smooth user experience.

[0035] The present invention optimizes the rendering and loading process of images in the virtual laboratory by introducing dynamic lighting calculation based on 3D models, cloud-based distributed rendering architecture and adaptive compression technology. In traditional technologies, the delay in image rendering and the jamming phenomenon in low-bandwidth environments seriously affect the students' interactive experience. The present invention significantly improves the rendering speed and image quality, reduces loading delays, and thus improves the overall performance of the virtual laboratory by dynamically adjusting the image compression ratio and optimizing the rendering strategy in real time.

[0036] The adaptive compression technology introduced in the present invention dynamically adjusts the image compression ratio according to device performance and network bandwidth, thereby reducing data transmission delay and bandwidth pressure while ensuring image quality. Compared with traditional static compression algorithms, the adaptive compression technology of the present invention can not only improve the loading efficiency in low-bandwidth environments, but also adjust the compression strategy according to the complexity and real-time changes of the experimental content, ensuring that each student can get a smooth virtual experiment experience in different network environments.

[0037] The present invention optimizes the distributed rendering architecture and reasonably distributes computing tasks to multiple computing nodes, thereby reducing the delay caused by insufficient computing resources in the rendering process. Under high load conditions, the system can intelligently balance the load to ensure the speed and quality of image rendering. Through the real-time feedback mechanism, the system can flexibly adjust the rendering strategy during student interaction to ensure that every operation in the experimental scene can receive a timely response.

[0038] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical scheme and inventive concept of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. A method for quickly generating a panoramic dynamic image online based on a 3D model, characterized in that: The steps include: S1, obtaining 3D model data and generating a 3D model through a 3D modeling tool; S2, simplifying the 3D model, and using the 3D model to obtain images from different viewing angles; S3, synthesizing the images of different viewing angles into a panoramic image by using an improved gradual transition algorithm, wherein the improved gradual transition algorithm performs a smooth transition of color and texture on the spliced ​​area; S4, performing real-time calculation of illumination, shadow, and reflection in the panoramic image, and updating the illumination effect of the panoramic image according to the illumination intensity; S5. Use particle system technology to generate dynamic effects for panoramic images; S6. Use the cloud distributed rendering architecture to distribute image generation tasks to multiple computing nodes, perform rendering processing through the cloud cluster, and use improved adaptive compression technology to dynamically adjust the image compression ratio in real time according to the user's network conditions and device performance; S7. According to the user's operation instruction, the rendering angle and dynamic effect of the panoramic image are adjusted in real time.

2. The method for quickly generating a panoramic dynamic image online based on a 3D model according to claim 1, characterized in that: The S2 specifically includes: S21. Use the QEM algorithm to simplify the mesh of the 3D model. The QEM algorithm calculates the geometric error of each triangular face. The geometric error is represented by an error matrix Q. The error matrix Q is the distance error between the vertex coordinates of each triangular face and the geometric center point of the adjacent face of the vertex coordinates: ； in, is the coordinate of each vertex on the triangle patch, is the geometric center coordinate of the triangle patch, and n is the number of vertices; The error matrix Q of each triangular face is applied to the simplification process. The triangular face with the smallest error is selected for simplification according to the error size. The geometric position of the simplified new triangular face is the weighted average of its error matrix Q. Iteratively update the simplified new triangle patches, and continue to select triangle patches with the smallest error for merging until the preset number of triangle patches M is reached; S22, simplifying the texture map of the 3D model, and obtaining a texture block size T based on the number of triangle patches M; S23. Evaluate the error between the simplified model and the original model using the error metric to obtain the error threshold , if the error is less than the preset threshold, the simplified model is judged to be qualified.

3. The method for quickly generating a panoramic dynamic image online based on a 3D model according to claim 1, characterized in that: The S3 specifically includes: S31, using an improved gradual transition algorithm to smoothly transition multiple perspective images, wherein the improved gradual transition algorithm uses a gradual factor G(x, y) for smoothing: ； Among them, d(x,y) represents the distance between the pixel point (x,y) in the stitching area and the stitching edge. is the maximum distance, C(x,y) represents the color value of the pixel (x,y), is the average color of the area, , is the regulating factor, is the natural exponential function; S32. Obtain a smoothed viewing angle image according to the gradient factor G(x,y): ； in, is the smoothed pixel value, is the pixel value of the previous image. is the pixel value of the image after stitching; S33, synthesizing the multiple perspective images into a panoramic image.

4. The method for quickly generating a panoramic dynamic image online based on a 3D model according to claim 1, characterized in that: The S4 specifically includes: S41, real-time calculation of the illumination intensity I(x,y) in the panoramic image: ； in, is ambient light, is diffuse reflected light, is the specular reflected light, To refract light, For scattered light; The diffusely reflected light The calculation is as follows: ； in, is the influence coefficient of the incident angle of light on diffuse reflection, is the dot product of the surface normal and the light source direction, is the intensity of the light source; The specular reflected light The calculation is as follows: ； in, is the mirror reflection coefficient, H is the half-distance vector, is the roughness index, V is the observer direction vector, which represents the unit vector pointing from the observer to the surface being rendered, is the dot product of the surface normal and the light source direction, is the dot product of the surface normal and the half-range vector, is the intensity of the light source; The refracted light The calculation formula is: ； in, is the Fresnel-Schlick approximation coefficient, is the refractive index of the medium, is the dot product of the surface normal and the light source direction, is the intensity of the light source; The scattered light The calculation formula is: ； in, and are the Rayleigh and Mie scattering coefficients, is the dot product of the surface normal and the light source direction, is the adjustment coefficient, is the intensity of the light source; S42: Update the lighting effect in the panoramic image according to the lighting intensity I(x, y) calculated in real time.

5. The method for quickly generating a panoramic dynamic image online based on a 3D model according to claim 1, characterized in that: The S5 specifically includes: S51. Apply particle system technology to generate dynamic effects in the panoramic image. The particle system is composed of a number of particles. The state of each particle is composed of position P(x, y, z), velocity V(x, y, z), acceleration A(x, y, z) and life cycle Definition: When generating particles, random distribution and gradient weighting methods are used. The initial position of the particles Randomization is performed based on the spatial distribution within the region, and when the particle position is updated, the Gaussian distribution is used to model the particle movement; S52. The update formula of the particle system is: ； ； in, is the new position of the particle, is the initial position of the particle, is the new velocity of the particle, is the initial velocity of the particle, is the time step; S53. According to the generated dynamic particle effect, the movement trajectory of the particles is adjusted in real time, and the generated dynamic effect is superimposed on the panoramic image.

6. The method for quickly generating a panoramic dynamic image online based on a 3D model according to claim 1, characterized in that: The S6 specifically includes: S61, using cloud-based distributed rendering architecture, distributes image rendering tasks to multiple computing nodes, and the rendering nodes dynamically allocate tasks based on the load balancing algorithm: ； in, is the computing task of the i-th rendering node, is the load of the i-th rendering node, is the computing power of the i-th rendering node, is the bandwidth, N is the total number of nodes, is the load of the ith node, The computing power of the ith node, is the bandwidth of the i-th node; S62. Obtain the rendering priority of each image block: ； in, is the priority of the image block, is the rendering time of the i-th image block, is the spatial complexity of the i-th image block, , is the regulating factor; S63. Adopt a rendering task transmission strategy based on multi-level network bandwidth control to dynamically adjust the bandwidth of data transmission: ； in, Bandwidth allocation for transmission tasks, is the maximum bandwidth, is the priority of the transmission task, M is the number of tasks; S64, providing real-time feedback and adjustment on the calculation results of each rendering node, calculating the workload of the rendering node through the feedback mechanism and adjusting the task allocation between the nodes, the feedback adjustment formula is: ； in, is the adjusted rendering task volume of the i-th rendering node, is the computing task of the i-th rendering node, is the global average load, is the load of the i-th rendering node, is the regulating factor; S65, using improved adaptive compression technology, dynamically adjust the image compression ratio in real time according to the user's network conditions and device performance to adapt to different bandwidths and device capabilities: ； in, is the compression ratio, is the transmission task bandwidth, is the priority of the image block, is the rendering task volume, is the node load, is the number of image blocks, is the image complexity, , is the adjustment coefficient, is the adaptive adjustment function, is the natural exponential function; S66, according to the compression ratio Compressing panoramic images; S67, combined with the quality evaluation of compressed images, optimize the adaptive compression algorithm, and the quality of compressed images for: ； in, is the entropy value of the image, For data transmission delay, is the maximum allowed delay, is the delay influence coefficient, is the complexity influence coefficient, is the image complexity.

7. A system for quickly generating panoramic dynamic images online based on a 3D model, executing a method for quickly generating panoramic dynamic images online based on a 3D model according to any one of claims 1 to 6, characterized in that: Includes the following modules: 3D model acquisition and processing module, used to collect 3D model data and simplify and optimize it; A panoramic image synthesis module is used to synthesize a panoramic image from multiple perspective images through cubic projection, and to calculate the overlapping part of the stitching area using image stitching technology; Dynamic lighting and effect generation module, used to calculate lighting, shadows, reflections and dynamic effects in panoramic images to generate natural dynamic effects; The rendering and distributed computing module distributes rendering tasks to multiple computing nodes through the cloud distributed rendering architecture, dynamically optimizes task allocation based on node load, bandwidth, and computing power, and performs parallel processing and optimization of rendering results. Adaptive compression and transmission optimization module, which is used to adjust the image compression ratio according to the user's network bandwidth and device performance, using adaptive compression technology to balance image quality and transmission efficiency, while reducing transmission delay through network optimization strategies; The real-time feedback and optimization module is used to provide feedback and adjust image details and resolution based on real-time rendering results.

Images