A regional image processing method and system for virtual exhibition hall display control

Abstract

The present invention discloses a regional image processing method and system for virtual exhibition hall display control, which relates to the field of image processing technology. The method includes: constructing a two-layer behavior model including a common behavior heat map and individual behavior portraits; establishing a three-dimensional spatial region from the current user's perspective based on the model, dividing it into a central region and a neighborhood region; formulating a thermal response strategy based on the common behavior heat map, making hierarchical decisions for the three-dimensional spatial region, and generating a benchmark image processing scheme; and executing the virtual exhibition hall display control in the three-dimensional space based on the scheme. This achieves the technical effect of improving the flexibility and rationality of resource allocation, and enhancing display efficiency and user experience.

Description

Technical Field

[0001] The present invention relates to the technical field of image processing, and in particular to a regional image processing method and system for virtual exhibition hall display control. Background Art

[0002] Existing virtual exhibition hall display technologies typically rely on fixed rendering strategies, displaying exhibition hall content through preset image processing schemes. These schemes are mainly based on hardware performance optimization or simple user interaction triggering mechanisms, such as switching scenes through mouse clicks or gesture swipes. However, as the scale of virtual exhibition halls expands and user behavior diversifies, existing technologies are gradually exposing limitations in resource allocation and user experience optimization. Technical issues such as lack of targeted resource allocation, waste of computing resources, and impact on display efficiency and user experience exist. Summary of the Invention

[0003] The present invention provides a regional image processing method and system for virtual exhibition hall display control, so as to solve the technical problems in the prior art of lack of targeted resource allocation, waste of computing resources, and impact on display efficiency and user experience, thereby achieving the technical effect of improving the flexibility and rationality of resource allocation, and improving display efficiency and user experience.

[0004] In a first aspect, the present invention provides a method for processing regional images for display control in a virtual exhibition hall, wherein the method for processing regional images for display control in a virtual exhibition hall comprises:

[0005] A two-layer behavior model based on user behavior is constructed, wherein the two-layer behavior model includes a common behavior heat map and an individual behavior portrait.

[0006] According to the two-layer behavior model, a three-dimensional space region under the current user's perspective is established, wherein the three-dimensional space region is divided into a central space region and a neighborhood space region.

[0007] A thermal response strategy is constructed, and a hierarchical processing decision is made for the three-dimensional space region in combination with the common behavior heat map, and the decision result is output as a reference area image processing solution.

[0008] Based on the reference area image processing solution, the display control of the virtual exhibition hall is carried out in combination with the three-dimensional space area.

[0009] In one feasible implementation, a two-layer behavior model based on user behavior is constructed, wherein the two-layer behavior model includes a common behavior heat map and an individual behavior profile, including:

[0010] Obtaining the gaze trajectory and operation behavior data of multiple users and analyzing them to obtain user interaction behavior data, wherein the user interaction behavior data includes gaze trajectory information and operation action information.

[0011] Based on the gaze trajectory information, the gaze trajectory distribution in the virtual exhibition hall area is obtained through statistical analysis, and the common behavior heat map is generated accordingly.

[0012] Based on the operation action information, the personality behavior portrait is constructed in combination with feature engineering.

[0013] Outputting the common behavior heat map and the individual behavior portrait is the double-layer behavior model.

[0014] In a feasible implementation, constructing the personality behavior profile based on the operation action information and combining feature engineering includes:

[0015] The operation action information is clustered and divided according to the unique identification identity of the user.

[0016] Perform feature extraction on the clustering results to obtain multiple groups of key behavioral feature data.

[0017] The perspective turning rate, action randomness entropy and gaze dwell time distribution corresponding to multiple groups of the key behavior feature data are calculated respectively, and the vectorized calculation results are the personality behavior portrait, wherein the personality behavior portrait is associated with a unique identification identity.

[0018] In a feasible implementation, based on the two-layer behavior model, a three-dimensional space region under the current user's perspective is established, wherein the three-dimensional space region is divided into a central space region and a neighborhood space region, including:

[0019] Based on the common behavior heat map and the current user perspective, the area with a heat value lower than the omitted heat limit is cropped to determine the central space area.

[0020] Based on the central space area, neighborhood growth is performed in combination with the personality behavior portrait to determine the corresponding neighborhood space area, wherein the neighborhood growth radius is calculated and determined based on the personality behavior portrait.

[0021] The central spatial region and the neighborhood spatial region are combined and smoothed.

[0022] In a feasible implementation, a thermal response strategy is constructed, and a hierarchical processing decision is made for the three-dimensional spatial region in combination with the common behavior heat map, and the decision result is output as a reference region image processing solution, including:

[0023] The thermal response strategy is constructed based on the processing performance parameters and hardware function information of the target scene, wherein the thermal response strategy includes multiple rendering sub-strategies, and each of the rendering sub-strategies has differentiated configurations in terms of image processing accuracy and computing resource scheduling.

[0024] The preset region semantic information is used in combination with the common behavior heat map to divide and mark the three-dimensional space region according to its interest level.

[0025] According to the mapping relationship in the thermal response strategy, the image rendering sub-strategy corresponding to the interest level is matched, and the matching result is output as the reference area image processing solution.

[0026] In a feasible implementation, based on the reference area image processing solution, display control of the virtual exhibition hall is performed in combination with the three-dimensional space area, including:

[0027] Object recognition is performed on the central spatial area and the neighborhood spatial area respectively to determine a resource calling target and a resource pre-storage target.

[0028] The reference area image processing solution is parsed to extract a first calling precision corresponding to the resource calling target and a second calling precision corresponding to the resource pre-stored target.

[0029] According to the resource calling target and the first calling precision, the corresponding real-time digital display resource is addressed and read to perform image rendering and display control.

[0030] Synchronously, according to the resource pre-storage target and the second call accuracy, the corresponding predicted digital display resource is read and stored in a cache medium.

[0031] In a feasible implementation, based on the reference area image processing solution, display control of the virtual exhibition hall is performed in combination with the three-dimensional space area, and then the method further includes:

[0032] If multiple users are detected at the same spatial location coordinates, the display area image corresponding to each user is extracted and a comparative analysis is performed.

[0033] Based on the comparison results of the display area images, invariant feature information is identified, and a shared image rendering context is constructed according to the invariant feature information, wherein the shared image rendering context is used for fast image reconstruction and image multiplexing among multiple users.

[0034] In a second aspect, the present invention further provides a regional image processing system for virtual exhibition hall display control, wherein the regional image processing system for virtual exhibition hall display control includes:

[0035] The two-layer behavior model construction module is used to construct a two-layer behavior model based on user behavior, wherein the two-layer behavior model includes a common behavior heat map and an individual behavior portrait.

[0036] The region establishment module is used to establish a three-dimensional space region under the current user's perspective based on the two-layer behavior model, wherein the three-dimensional space region is divided into a central space region and a neighborhood space region.

[0037] The image processing decision module is used to construct a thermal response strategy, and make a hierarchical processing decision for the three-dimensional space area in combination with the common behavior thermal map, and output the decision result as a reference area image processing solution.

[0038] The display control module is used to control the display of the virtual exhibition hall based on the reference area image processing solution and in combination with the three-dimensional space area.

[0039] The present invention discloses a regional image processing method and system for display control of a virtual exhibition hall, comprising: constructing a two-layer behavior model based on user behavior, the model including a common behavior heat map and an individual behavior portrait; on this basis, establishing a three-dimensional space region from the current user's perspective according to the two-layer behavior model, and dividing the space region into a central space region and a neighborhood space region; subsequently, constructing a thermal response strategy, and performing hierarchical processing decisions on the three-dimensional space region in combination with the common behavior heat map, and outputting a reference region image processing scheme; finally, based on the image processing scheme, implementing display control of the virtual exhibition hall within the three-dimensional space region. The regional image processing method and system for display control of a virtual exhibition hall disclosed by the present invention solve the technical problems of lack of pertinence in resource allocation, waste of computing resources, and impact on display efficiency and user experience, and achieve the technical effect of improving the flexibility and rationality of resource allocation, and improving display efficiency and user experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 The figure is a flow chart of a method for regional image processing for display control of a virtual exhibition hall according to the present invention.

[0041] Figure 2 The figure is a structural diagram of a regional image processing system for display control of a virtual exhibition hall according to the present invention.

[0042] Explanation of the accompanying drawings: two-layer behavior model construction module 11, area establishment module 12, image processing decision module 13, display control module 14. DETAILED DESCRIPTION

[0043] The above technical solution will be described in detail below in conjunction with the accompanying drawings and specific implementation methods of the specification to better understand the above technical solution. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments of the present invention. It should be understood that the present invention is not limited to the example embodiments used only to explain the present invention. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention. In addition, it should be noted that, for the convenience of description, only the parts related to the present invention, rather than all, are shown in the drawings.

[0044] Example 1, as Figure 1 The flowchart of a method for processing regional images for controlling display in a virtual exhibition hall according to the present invention is shown below. The method for processing regional images for controlling display in a virtual exhibition hall includes:

[0045] A two-layer behavior model based on user behavior is constructed, wherein the two-layer behavior model includes a common behavior heat map and an individual behavior portrait.

[0046] Specifically, the two-layer behavior model contains two levels of behavior analysis models (common behavior heat map and individual behavior portrait), which is used to capture the common behavior of user groups and the individual behavior of individual users at the same time. Among them, the common behavior heat map is generated by statistical analysis of the behavior data of multiple users, which is used to reflect the centralized trend and hot spots of the behavior of user groups in the virtual exhibition hall. Preferably, the frequency or intensity of the behavior is usually represented by the depth of color. The individual behavior portrait is the modeling result of the behavioral characteristics of a single user. By analyzing the user's operation action information, it can extract feature data reflecting the user's unique behavior pattern, such as the viewing angle turning rate, the action randomness entropy value and the distribution of the gaze dwell time, and vectorize it as a personal behavior portrait.

[0047] In the above steps, the common behavior heat map helps the system identify areas of common interest among user groups, prioritizing these areas when allocating resources, avoiding blind allocation and waste of resources. Furthermore, individual behavior profiling enables the system to personalize display control based on each user's specific behavioral characteristics, further enhancing the user experience. This combined common and individual behavior modeling approach improves display effectiveness and, in turn, optimizes the overall performance of the virtual exhibition hall.

[0048] In some embodiments, a two-layer behavior model based on user behavior is constructed, wherein the two-layer behavior model includes a common behavior heat map and an individual behavior profile, including:

[0049] Obtaining the gaze trajectory and operation behavior data of multiple users and analyzing them to obtain user interaction behavior data, wherein the user interaction behavior data includes gaze trajectory information and operation action information; based on the gaze trajectory information, statistically analyzing to obtain the gaze trajectory distribution of the virtual exhibition hall area, and correspondingly generating the common behavior heat map; based on the operation action information, combining feature engineering to construct the individual behavior portrait; outputting the common behavior heat map and the individual behavior portrait as the two-layer behavior model.

[0050] Specifically, user interaction behavior data refers to the various data generated when users interact with the system in the virtual exhibition hall, including historical user gaze trajectory information and operational action information. Gaze trajectory information refers to the path and location of the user's gaze movement in the virtual exhibition hall (represented as a time-series data sequence of gaze direction vectors), reflecting the user's attention to different areas. Operational action information includes various user operations in the virtual exhibition hall, such as clicking, sliding, and zooming. These operations can reflect the user's interests and interaction habits.

[0051] Specifically, eye tracking devices or human-computer interaction devices record the user's gaze trajectory and, through system logs, record user actions such as clicks and swipes. For example, in a virtual art gallery, this data can include the time a user spends in front of a painting, the path of their gaze, and whether they click to enlarge the painting. Together, these data constitute user interaction behavior data, providing foundational data support for subsequent model construction.

[0052] Specifically, the collected gaze trajectory information is then statistically analyzed. That is, by calculating the dwell frequency and cumulative dwell time of all gazes in different areas, a common behavior heat map is generated. For example, in a virtual exhibition hall, if most users' gazes are focused on certain paintings, these areas will appear as high-heat areas on the heat map. These areas can be considered the focus of general attention of the user group and may require higher attention or image processing capabilities. The above-mentioned common behavior heat map can intuitively reflect the common behavioral characteristics of the user group, providing an important basis for subsequent resource allocation and display control.

[0053] Specifically, by extracting and converting the features of the operation action information, key data that effectively describe the user's behavior characteristics can be extracted, such as the viewing angle turning rate, the action randomness entropy value, and the distribution of the gaze dwell time; then, these features are vectorized to form a personality behavior portrait, in which each feature indicator corresponds to a vector dimension, that is, the personality behavior portrait is represented as a multidimensional vector. For example, for a certain user, his personality behavior portrait may show that he prefers to quickly browse multiple exhibits and has a higher frequency of attention to certain specific types of exhibits. The personality behavior portrait obtained through the above process can accurately reflect the personalized characteristics of individual users such as habits and preferences, which helps to provide each user with a personalized display experience.

[0054] Furthermore, the generated common behavior heatmaps and individual behavior profiles are integrated into a two-tier behavior model, which serves as the basis for subsequent display control and resource allocation decisions. This two-tier behavior model not only helps improve the flexibility and rationality of resource allocation, but also helps enhance display effectiveness and user experience.

[0055] In some implementations, constructing the personality behavior profile based on the operation action information in combination with feature engineering includes:

[0056] According to the user's unique identification identity, the operation action information is clustered; feature extraction is performed on the clustering results to obtain multiple groups of key behavior feature data; the perspective turning rate, action randomness entropy value and gaze residence time distribution corresponding to the multiple groups of key behavior feature data are calculated respectively, and the vectorized calculation result is the personality behavior portrait, wherein the personality behavior portrait is associated with a unique identification identity.

[0057] Specifically, first, the operation action information generated by a specific user in the system is extracted, and the user's operation action information is grouped according to the user identity based on the unique identification identity bound to it, ensuring that the behavior data of each user is processed independently; then, feature extraction is performed on each type of operation sample after clustering, and multiple groups of key behavior feature data are constructed, such as the viewing angle and frequency sequence, gesture trajectory curvature, operation start and end time difference, the number of turning points in the eye movement path, etc. In other words, key behavior feature data can be considered as the basic representation of user personality.

[0058] Furthermore, based on the above key behavioral feature data, the perspective turning rate (average and peak values ​​of angular velocity changes) corresponding to each set of features is calculated; the action randomness entropy of the behavioral sequence is calculated (such as Shannon entropy and sample entropy, which measure the regularity of operations); and the distribution of gaze dwell time (average gaze time and dwell frequency on different areas or interface elements) is statistically analyzed. The above multiple indicators can be considered as a more comprehensive representation of user personality based on key behavioral feature data.

[0059] Furthermore, the extracted key behavioral feature data are normalized and quantified to form a behavioral feature vector with a fixed dimension as a personality behavior portrait to facilitate subsequent model construction and analysis. The personality behavior portrait is associated with a unique identification tag for subsequent display control and resource allocation.

[0060] Through the above process, a personalized behavior profile is constructed based on operational information, enabling refined and personalized analysis of user behavior and providing data support for subsequent personalized display. Specifically, the personalized behavior profile is formed by quantizing key behavioral feature data, effectively reflecting user operating habits and preferences, thereby achieving precise resource allocation and optimized display strategies.

[0061] According to the two-layer behavior model, a three-dimensional space region under the current user's perspective is established, wherein the three-dimensional space region is divided into a central space region and a neighborhood space region.

[0062] Specifically, the current user perspective refers to the user's observation or operation position and direction at a specific interaction time node, which is determined by perspective coordinates (such as camera position + orientation vector), eye tracking or interface focus information; the three-dimensional spatial area is a spatial area structure constructed around the user's current perspective, which is used to map the user's attention intensity or behavioral response possibility to different spatial positions in the environmental scene, and guide the scope that needs to be paid attention to in subsequent image processing.

[0063] Specifically, the three-dimensional space area includes the central space area and the neighborhood space area. The central space area corresponds to the high-attention space area near the user's line of sight or the operation focus; the neighborhood space area is the low-attention area distributed around the central space area, indicating the environmental space range that may be noticed or indirectly affected; through dynamic modeling of the neighborhood space, it helps to prepare in advance for the user's possible behavioral trends such as turning and attention switching, thereby enhancing the forward-looking response and personalized adaptability of the interactive system.

[0064] In some embodiments, based on the two-layer behavior model, a three-dimensional space region under the current user's perspective is established, wherein the three-dimensional space region is divided into a central space region and a neighborhood space region, including:

[0065] Based on the common behavior heat map and the current user perspective, the area with a heat value lower than the omitted heat limit is cropped to determine the central space area; based on the central space area, neighborhood growth is performed in combination with the individual behavior portrait to determine the corresponding neighborhood space area, wherein the neighborhood growth radius is calculated and determined based on the individual behavior portrait; the central space area and the neighborhood space area are combined and smoothed.

[0066] Specifically, the thermal value refers to the numerical value that represents the frequency or intensity of behavior in the heat map; the omitted thermal limit refers to a set threshold used to determine which areas have low thermal values ​​and can be ignored.

[0067] Specifically, the system first identifies the core area of ​​the user's current attention using a common behavior heat map and the user's current perspective. For example, in a virtual exhibition hall, the system uses the user's current perspective position and direction, combined with the common behavior heat map, to crop out areas with heat values ​​below a set threshold (i.e., omitting the heat limit) (such as pure black or fixed background areas), retaining the high-heat areas of the user's attention as the central spatial area. This cropping ensures that resources are focused on the areas the user truly cares about, avoiding unnecessary computation for areas of no interest.

[0068] Specifically, the algorithm then analyzes the user's attention shifting tendency and the range of their behavior, taking the central spatial region as the core and combining it with the user's behavioral profile. It then dynamically calculates the neighborhood growth radius and performs a neighborhood expansion operation in three-dimensional space. For example, a greater viewing angle turn rate, a greater entropy value for randomness, and a more dispersed distribution of gaze dwell time indicate greater randomness in the user's behavior, corresponding to a larger growth radius. For example, if a user prefers quick browsing, the neighborhood growth radius is appropriately expanded to include more content of potential interest.

[0069] Furthermore, the central spatial area and the neighborhood spatial area are merged and smoothed using an interpolation algorithm or boundary smoothing technology to ensure a natural transition between the boundary of the central area and the neighborhood area and avoid an abrupt boundary feeling. The above processing not only improves the visual effect, but also makes the user's browsing experience smoother.

[0070] The above process establishes three-dimensional spatial regions based on a two-layer behavioral model, significantly improving the flexibility and rationality of resource allocation. First, low-heat value areas are cut off to ensure that resources are concentrated in areas of real user interest, avoiding resource waste. The neighborhood growth process dynamically adjusts the neighborhood range based on individual behavioral profiles, making the displayed content more personalized to the user's needs. Combining and smoothing ensures a natural transition between region boundaries, enhancing the fluidity and aesthetics of the display. This user behavior-based regional division method not only helps improve display efficiency, but also significantly enhances the user experience.

[0071] A thermal response strategy is constructed, and a hierarchical processing decision is made for the three-dimensional space region in combination with the common behavior heat map, and the decision result is output as a reference area image processing solution.

[0072] Specifically, the thermal response strategy is a dynamic resource allocation strategy based on user behavior heatmaps. It allocates different processing resources and rendering quality based on the importance of different spatial regions, achieving a balance between processing efficiency and user experience. Three-dimensional spatial regions can be prioritized based on their level of interest or importance, and corresponding image processing solutions can be assigned to each level.

[0073] Specifically, the reference area image processing scheme is the final image processing scheme output according to the hierarchical processing decision, which is used to guide the display control of the virtual exhibition hall, including multiple hierarchical area units and corresponding display control methods, control parameters, etc.

[0074] In some embodiments, a thermal response strategy is constructed, and a hierarchical processing decision is made for the three-dimensional spatial region in combination with the common behavior heat map, and the decision result is output as a reference region image processing solution, including:

[0075] The thermal response strategy is constructed based on the processing performance parameters and hardware function information of the target scene, wherein the thermal response strategy includes multiple rendering sub-strategies, and each rendering sub-strategy has a differentiated configuration in terms of image processing accuracy and computing resource scheduling; using preset regional semantic information and combining with the common behavior heat map, the three-dimensional space area is divided into interest levels and marked; according to the mapping relationship in the thermal response strategy, the image rendering sub-strategy corresponding to the interest level is matched, and the matching result is output as the reference area image processing solution.

[0076] Specifically, the rendering sub-strategy is a subset of the thermal response strategy, with different configurations for image rendering accuracy, computational complexity, resource usage, and other parameters. High-quality rendering is detailed but resource-intensive, such as full rendering with high resolution, high frame rate, high textures, and high shadows. Medium-quality rendering balances performance and accuracy, such as low-quality rendering (fast response but compressed detail).

[0077] Specifically, the regional semantic information is used to describe the semantic attributes of each sub-region in the three-dimensional space, such as "operation area", "observation area", "edge area", etc., providing a semantic basis for interest level classification.

[0078] Exemplarily, a high-quality rendering sub-strategy uses the original high-resolution image or an equivalent high-precision image source; retains complete details such as texture, shadow, and reflection, and is suitable for central high-interest areas where users gaze for long periods of time and interact frequently; it is suitable for scenarios with sufficient terminal computing power, such as: using high-throughput graphics card renderers, virtual reality headsets, etc.; a medium-quality rendering sub-strategy needs to comprehensively balance processing efficiency and visual fidelity, such as using new neural network processing methods such as AI super-resolution algorithms and AI interpolation models (based on CUDA or tensor cores) to obtain approximate high-quality effects with medium image resource input; among them, this strategy can be used for some medium-interest areas, and is particularly suitable for optimization processing in AI accelerated hardware environments; the low-quality rendering sub-strategy is mainly suitable for edge or low-attention areas, and can use traditional image enhancement methods such as interpolation methods based on optical flow estimation, edge anti-aliasing algorithms, etc. for fast compressed rendering, which can effectively reduce processing load and improve system response speed, while supporting the operation of low-specification hardware and resource-sensitive devices.

[0079] Specifically, we first collect the image processing performance parameters (such as frame rate upper limit, GPU computing power, cache bandwidth, etc.) and terminal hardware functions (such as graphics card model, parallel processing capability) of the target usage scenario; combine the above parameters to establish a responsiveness model, and construct a thermal response strategy containing multiple rendering sub-strategies, each sub-strategy has a different image processing level (such as resolution level, detail level, mapping frequency, etc.); among them, the thermal response strategy has a strategy-resource mapping capability, which can match the interest area and resource allocation mode on demand; then, for the three-dimensional space area constructed in the previous stage, the regional semantic information (such as identifying "user's permanent perspective" and "frequently operated objects") and the thermal value in the common behavior heat map are integrated to perform weight calculation (such as weighted sum), and the interest level division result is output. The interest level division result is used to guide differentiated image rendering processing.

[0080] Furthermore, based on the level-strategy mapping table in the thermal response strategy, it automatically matches rendering sub-strategies to different interest levels. The matching results of all sub-strategies are combined to generate a baseline region image processing plan, which is used to indicate the application of different image processing methods in different spatial locations, achieving optimal resource allocation and maximizing the user experience. This allows the system to minimize overall image computation overhead without sacrificing clarity in key visual areas.

[0081] Based on the reference area image processing solution, the display control of the virtual exhibition hall is carried out in combination with the three-dimensional space area.

[0082] In some embodiments, based on the reference area image processing solution, display control of the virtual exhibition hall is performed in combination with the three-dimensional space area, including:

[0083] Object recognition is performed on the central spatial area and the neighborhood spatial area respectively to determine a resource call target and a resource pre-stored target; the reference area image processing scheme is parsed to extract a first call accuracy corresponding to the resource call target and a second call accuracy corresponding to the resource pre-stored target; according to the resource call target and the first call accuracy, the corresponding instant digital display resource is addressed and read for image rendering and display control; synchronously, according to the resource pre-stored target and the second call accuracy, the corresponding predicted digital display resource is read and stored in a cache medium.

[0084] Specifically, first, object recognition is performed on the central space area and the neighborhood space area respectively. Based on the regional perspective and interaction frequency analysis, the instant interactive objects that need to be rendered currently (defined as resource call targets) are automatically identified, and the candidate objects that will enter the user's attention range (defined as resource pre-storage targets) are predicted; in other words, the resource call targets are areas or objects that need to be rendered and displayed immediately, including exhibits, booths, lights, walls, etc. in the virtual exhibition hall, and the resource pre-storage targets refer to areas or objects that may need to be rendered in the future, which are used for preloading to improve display efficiency. For example, in a virtual exhibition hall, the main exhibits in the central space area are identified as resource call targets, while the secondary exhibits or backgrounds in the neighborhood space area are identified as resource pre-storage targets. The above-mentioned identification process can ensure that high-priority content is given priority, and at the same time helps to preload content that may be needed.

[0085] Specifically, the rendering sub-strategies matched by each area in the benchmark area image processing scheme are parsed, and the first call precision corresponding to the resource call target and the second call precision corresponding to the resource pre-stored target are extracted therefrom; wherein the above-mentioned call precisions are matched with parameters such as image resolution, texture level, model detail level, anti-aliasing strength, etc. in the rendering sub-strategies.

[0086] Specifically, instant digital display resources refer to the target digital resources that need to be used immediately, and need to meet the first call accuracy, such as high-precision models, videos, etc.; predicted digital display resources refer to digital resources that may be needed based on user behavior predictions, which are usually pre-loaded into the cache and meet the second call accuracy.

[0087] Furthermore, the graphics rendering pipeline (such as graphics interfaces based on OpenGL / DirectX / Vulkan) is called to perform real-time image rendering based on the acquired instant digital display resources, and perform dynamic display control to ensure that the core content that the user is currently concerned about is displayed with the highest quality; at the same time, according to the resource pre-storage target and the second call accuracy, the resources of the neighboring spatial area (i.e., the predicted digital display resources) are read from the storage device and pre-loaded into the local cache medium (such as GPU video memory, system cache, VRAM buffer pool, etc.) to reduce the delay when accessing these resources in the future and improve the display efficiency.

[0088] Through the above process, a dynamic display resource scheduling mechanism driven by image processing strategies can be realized in a 3D virtual exhibition hall, thereby improving the overall resource utilization efficiency, image rendering response speed and scalability of the display system while ensuring a highly immersive user experience.

[0089] In some implementations, based on the reference area image processing solution, display control of the virtual exhibition hall is performed in combination with the three-dimensional space area, and then the method further includes:

[0090] If multiple users are detected at the same spatial position coordinates, the display area images corresponding to each user are extracted and a comparative analysis is performed; based on the comparison results of the display area images, invariant feature information is identified, and based on the invariant feature information, a shared image rendering context is constructed, wherein the shared image rendering context is used for fast image reconstruction and image reuse among multiple users.

[0091] Specifically, the display area image refers to the image content seen by the user from the current perspective in the virtual exhibition hall; invariant feature information refers to the visual features that remain unchanged in the display area images of multiple users. The shared image rendering context is a shared rendering environment used to store and manage image resources and rendering parameters that can be reused by multiple users.

[0092] Exemplarily, the invariant feature information includes motion direction vectors, material texture features, and lighting parameter features. The motion direction vector refers to the direction and speed of movement of an object in an image and is used for dynamic scene analysis. Material texture features refer to the surface material and texture information of an object in an image, such as color and roughness. Lighting parameter features refer to the lighting conditions in an image, such as the direction and intensity of the light source.

[0093] Optionally, if multiple users are detected at the same or approximate spatial position coordinates through the head-mounted device posture data or behavior prediction trajectory, the shared rendering logic is triggered to extract the display area image data of each user's current perspective, including image frame cache information, rendering parameter set, etc.; then, the display area images are compared to identify similar fragments or stable presentation content (such as fixed exhibits and their materials, textures, shadows, etc.), and image reconstruction difference analysis is performed. The obtained invariant feature information indicates the same aspects of the displays of different users and can be reused.

[0094] Optionally, a shared image rendering context is constructed based on the aforementioned invariant feature information, allowing multiple users to reuse images within a shared area. The shared image rendering context may include a unified scene graph structure, a pre-rendered frame buffer, a common material buffer, and image reconstruction parameter configuration. For example, invariant material texture features and lighting parameter features can be stored in the shared rendering context so that they can be shared by multiple users, thereby reducing duplicate computation and wasted resources.

[0095] By identifying invariant feature information and constructing a shared image rendering context, the above process significantly improves the flexibility and rationality of resource allocation. The comparative analysis and identification of invariant feature information ensure that sharing possibilities between multiple users are fully utilized, reducing repeated computations. The construction of a shared image rendering context enables efficient resource reuse across multiple users, avoiding resource waste. Rapid image reconstruction and reuse significantly improve display efficiency, especially in multi-user scenarios, enhancing the overall system performance and user experience, and enabling efficient and personalized display control in virtual exhibition halls.

[0096] In summary, the regional image processing method for virtual exhibition hall display control provided by the present invention has the following technical effects:

[0097] By constructing a two-layer behavior model based on user behavior, which includes a common behavior heat map and an individual behavior portrait; on this basis, a three-dimensional spatial area is established from the current user's perspective according to the two-layer behavior model, and the spatial area is divided into a central spatial area and a neighborhood spatial area; then, a thermal response strategy is constructed, and a hierarchical processing decision is made on the three-dimensional spatial area in combination with the common behavior heat map, and the output result is a benchmark area image processing scheme; finally, based on the image processing scheme, the display control of the virtual exhibition hall is implemented in the three-dimensional spatial area, thereby achieving the technical effect of improving the flexibility and rationality of resource allocation, and improving the display efficiency and user experience.

[0098] Example 2, as Figure 2 This is a structural diagram of a regional image processing system for virtual exhibition hall display control according to the present invention. For example, Figure 1The flow chart of the regional image processing method for virtual exhibition hall display control in the present invention can be shown as follows: Figure 2 The structure shown is implemented.

[0099] Based on the same concept as the regional image processing method for virtual exhibition hall display control in the above embodiment, the present invention also provides a regional image processing system for virtual exhibition hall display control, comprising:

[0100] The two-layer behavior model construction module 11 is used to construct a two-layer behavior model based on user behavior, wherein the two-layer behavior model includes a common behavior heat map and an individual behavior portrait.

[0101] The region establishing module 12 is configured to establish a three-dimensional space region under the current user's perspective according to the two-layer behavior model, wherein the three-dimensional space region is divided into a central space region and a neighborhood space region.

[0102] The image processing decision module 13 is used to construct a thermal response strategy, and make a hierarchical processing decision for the three-dimensional space area in combination with the common behavior thermal map, and output the decision result as a reference area image processing solution.

[0103] The display control module 14 is configured to control the display of the virtual exhibition hall based on the reference area image processing solution and in combination with the three-dimensional space area.

[0104] In some embodiments, the two-layer behavior model building module 11 includes:

[0105] The user interaction behavior data acquisition unit is used to acquire the sight track and operation behavior data of multiple users and analyze and acquire the user interaction behavior data, wherein the user interaction behavior data includes sight track information and operation action information.

[0106] The common behavior heat map generating unit is used to obtain the sight track distribution of the virtual exhibition hall area through statistical analysis based on the sight track information, and generate the common behavior heat map accordingly.

[0107] A personality behavior portrait construction unit is used to construct the personality behavior portrait based on the operation action information in combination with feature engineering.

[0108] The double-layer behavior model output unit is used to output the common behavior heat map and the individual behavior portrait as the double-layer behavior model.

[0109] In some implementations, the personality behavior portrait construction unit in the two-layer behavior model construction module 11 includes:

[0110] The operation action information clustering unit is used to cluster the operation action information according to the unique identification identity of the user.

[0111] The key behavior feature data extraction unit is used to perform feature extraction on the clustering results to obtain multiple groups of key behavior feature data.

[0112] The personality behavior portrait quantization unit is used to respectively calculate the viewing angle turning rate, action randomness entropy value and gaze dwell time distribution corresponding to multiple groups of the key behavior feature data, and vectorize the calculation result into the personality behavior portrait, wherein the personality behavior portrait is associated with a unique identification identity.

[0113] In some embodiments, the region establishing module 12 includes:

[0114] The central space area determination unit is used to cut out the area with a heat value lower than the omitted heat limit based on the common behavior heat map and the current user perspective to determine the central space area.

[0115] A neighborhood space area determination unit is used to perform neighborhood growth based on the central space area and in combination with the personality behavior portrait to determine the corresponding neighborhood space area, wherein the neighborhood growth radius is calculated and determined based on the personality behavior portrait.

[0116] The three-dimensional spatial region combining and smoothing unit is used to combine and smooth the central spatial region and the neighborhood spatial region.

[0117] In some embodiments, the image processing decision module 13 includes:

[0118] A thermal response strategy construction unit is used to construct the thermal response strategy based on the processing performance parameters and hardware function information of the target scene, wherein the thermal response strategy includes multiple rendering sub-strategies, and each rendering sub-strategy has a differentiated configuration in terms of image processing accuracy and computing resource scheduling.

[0119] The three-dimensional space region interest level division unit is used to divide and mark the three-dimensional space region by interest level using preset region semantic information in combination with the common behavior heat map.

[0120] The reference area image processing solution generating unit is used to match the image rendering sub-strategy corresponding to the interest level according to the mapping relationship in the thermal response strategy, and output the matching result as the reference area image processing solution.

[0121] In some embodiments, the display control module 14 includes:

[0122] The object identification and resource target determination unit is used to perform object identification on the central space area and the neighborhood space area respectively, and determine the resource calling target and the resource pre-storage target.

[0123] The reference area image processing solution parsing unit is used to parse the reference area image processing solution and extract the first calling precision corresponding to the resource calling target and the second calling precision corresponding to the resource pre-stored target.

[0124] The real-time digital display resource calling unit is used to address and read the corresponding real-time digital display resource to perform image rendering and display control according to the resource calling target and the first calling accuracy.

[0125] The predicted digital display resource pre-storage unit is used to synchronously read the corresponding predicted digital display resource according to the resource pre-storage target and the second call accuracy and store it in a cache medium.

[0126] In some embodiments, the system further comprises:

[0127] The multi-user image comparison and analysis unit is used to extract the display area image corresponding to each user and perform comparison analysis if it is detected that multiple users are located at the same spatial position coordinates.

[0128] A sharing and image multiplexing unit is used to identify invariant feature information based on the comparison results of the display area image, and to construct a shared image rendering context based on the invariant feature information, wherein the shared image rendering context is used for fast image reconstruction and image multiplexing between multiple users.

[0129] It should be understood that the embodiments mentioned in this specification focus on their differences from other embodiments. The specific embodiments in the aforementioned embodiment one are also applicable to the regional image processing system for virtual exhibition hall display control described in embodiment two. For the sake of brevity of the specification, no further elaboration will be given here.

[0130] It should be understood that the embodiments disclosed in the present invention and the above description can enable those skilled in the art to use the present invention to implement the present invention. At the same time, the present invention is not limited to the embodiments mentioned above. It should be understood that those skilled in the art can still modify the technical solutions described in the above embodiments or replace some of the technical features therein with equivalents; and such modifications or replacements do not deviate from the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention and are all included in the scope of protection of the present invention.

Claims

1. A regional image processing method for virtual exhibition hall display control, characterized in that: The method comprises: Obtaining the gaze trajectory and operation behavior data of multiple users and analyzing and obtaining user interaction behavior data, wherein the user interaction behavior data includes gaze trajectory information and operation action information; Based on the gaze trajectory information, statistical analysis is performed to obtain gaze trajectory distribution in the virtual exhibition hall area, and a common behavior heat map is generated accordingly; Based on the operation action information, a personality behavior profile is constructed in combination with feature engineering; Constructing a two-layer behavior model based on the common behavior heat map and the individual behavior portrait; According to the common behavior heat map of the two-layer behavior model and the current user perspective, the area with a heat value lower than the omitted heat limit is cut off to determine the central space area; Based on the central spatial area, neighborhood growth is performed in combination with the personality behavior portrait to determine a corresponding neighborhood spatial area, wherein the neighborhood growth radius is calculated and determined based on the personality behavior portrait; Combining and smoothing the central spatial region and the neighborhood spatial region to establish a three-dimensional spatial region under the current user's perspective; Constructing a thermal response strategy, and making a hierarchical processing decision for the three-dimensional space region in combination with the common behavior heat map, and outputting the decision result as a reference region image processing solution; Based on the reference area image processing solution, the display control of the virtual exhibition hall is carried out in combination with the three-dimensional space area.

2. The method for regional image processing for virtual exhibition hall display control according to claim 1, characterized in that: Based on the operation action information, the personality behavior profile is constructed in combination with feature engineering, including: Clustering the operation action information according to the user's unique identification identity; Perform feature extraction on the clustering results to obtain multiple sets of key behavioral feature data; The perspective turning rate, action randomness entropy and gaze dwell time distribution corresponding to multiple groups of the key behavior feature data are calculated respectively, and the vectorized calculation results are the personality behavior portrait, wherein the personality behavior portrait is associated with a unique identification identity.

3. The method for regional image processing for virtual exhibition hall display control according to claim 2, characterized in that: Constructing a thermal response strategy, and combining the common behavior heat map to make a hierarchical processing decision for the three-dimensional space area, outputting the decision result as a reference area image processing solution, including: Constructing the thermal response strategy based on the processing performance parameters and hardware function information of the target scene, wherein the thermal response strategy includes multiple rendering sub-strategies, and each rendering sub-strategy has a differentiated configuration in terms of image processing accuracy and computing resource scheduling; Using the preset area semantic information and combining it with the common behavior heat map, the three-dimensional space area is divided into interest levels and marked; According to the mapping relationship in the thermal response strategy, the image rendering sub-strategy corresponding to the interest level is matched, and the matching result is output as the reference area image processing solution.

4. The method for regional image processing for virtual exhibition hall display control according to claim 3, characterized in that: Based on the reference area image processing solution, display control of the virtual exhibition hall is performed in combination with the three-dimensional space area, including: Performing object recognition on the central spatial area and the neighborhood spatial area respectively to determine a resource calling target and a resource pre-storage target; Parsing the reference area image processing solution, extracting a first call accuracy corresponding to the resource call target and a second call accuracy corresponding to the resource pre-stored target; According to the resource call target and the first call precision, addressing and reading the corresponding real-time digital display resource to perform image rendering and display control; Synchronously, according to the resource pre-storage target and the second call accuracy, the corresponding predicted digital display resource is read and stored in a cache medium.

5. The method for regional image processing for virtual exhibition hall display control according to claim 1, characterized in that: Based on the reference area image processing solution, the display control of the virtual exhibition hall is performed in combination with the three-dimensional space area, and then the method further includes: If multiple users are detected at the same spatial location coordinates, the display area image corresponding to each user is extracted and a comparative analysis is performed; Based on the comparison results of the display area images, invariant feature information is identified, and a shared image rendering context is constructed according to the invariant feature information, wherein the shared image rendering context is used for fast image reconstruction and image multiplexing among multiple users.

6. A regional image processing system for virtual exhibition hall display control, characterized in that: A regional image processing method for controlling display of a virtual exhibition hall, for executing any one of claims 1 to 5, comprising: A two-layer behavior model construction module is used to construct a two-layer behavior model based on user behavior, wherein the two-layer behavior model includes a common behavior heat map and an individual behavior portrait; A region establishment module, configured to establish a three-dimensional spatial region under the current user's perspective based on the two-layer behavior model, wherein the three-dimensional spatial region is divided into a central spatial region and a neighborhood spatial region; An image processing decision module is used to construct a thermal response strategy, and to make a hierarchical processing decision for the three-dimensional space region in combination with the common behavior thermal map, and output the decision result as a reference region image processing solution; The display control module is used to control the display of the virtual exhibition hall based on the reference area image processing solution and in combination with the three-dimensional space area.

Images