Garden design method based on garden simulation

Abstract

The invention relates to a garden design method based on garden simulation. The method comprises the following steps: generating a sightseeing path containing space coordinates, sight direction and stay duration based on tourist dynamic behavior data, generating a continuous video sequence in combination with a dynamic field chain algorithm, and constructing an environmental acoustic model by using a ray tracing method to simulate a sound pressure level attenuation curve. A visual rhythm curve and a sound scene rhythm curve are extracted, a dynamic time warping algorithm is adopted to quantify the matching degree, path parameters are automatically and iteratively adjusted until audio-visual cooperation reaches the standard, and finally time-space aligned multi-sensory space sequence data are generated and mapped to a virtual reality platform. The problems of seeing and hearing simulation splitting, low acoustic model precision and poor efficiency depending on manual adjustment in traditional garden design are solved, high-precision co-simulation and automatic design optimization of multi-sensory experience in a dynamic touring scene are realized, and the authenticity of immersive experience and the scheme generation efficiency are remarkably improved.

Description

Technical Field

[0001] The present invention belongs to the cross - field of computer - aided design and landscape engineering, and particularly relates to a landscape design method based on landscape simulation. Background Art

[0002] In the landscape design method based on landscape simulation, the core of the technical problem lies in how to accurately construct the visual and auditory perception space sequence to simulate the dynamic experience of tourists in the landscape. This process involves the complex interaction of path planning, visual sequence generation, and environmental acoustic simulation in the three - dimensional model.

[0003] Traditional methods rely on manually presetting the tour path, lacking data - driven analysis of tourists' dynamic behaviors, resulting in a high path overlap rate, a low visual framing density, and an inability to quantify the coherence of the visual rhythm.

[0004] At the same time, the acoustic simulation of open landscapes mostly uses simplified geometric acoustic models, ignoring the non - linear characteristics of material reflection and sound source distance attenuation, resulting in a large error in the matching of sound pressure level and visual focus, and the fragmentation of multi - sensory experience.

[0005] In addition, the adjustment of visual and acoustic parameters relies on manual trial - and - error, lacking cross - modal collaborative algorithms, with a long design iteration cycle and difficulty in meeting the requirements of audio - visual synchronization.

[0006] Therefore, there is an urgent need for a landscape design method that integrates dynamic behavior data, high - precision acoustic modeling, and audio - visual collaborative optimization to improve the authenticity of multi - sensory experience and design efficiency. Summary of the Invention

[0007] Based on this, it is necessary to provide a landscape design method based on landscape simulation for the above - mentioned technical problems.

[0008] In a first aspect, the present application provides a landscape design method based on landscape simulation, including:

[0009] S1: In the three - dimensional model of the landscape, generate multiple tour paths based on tourists' dynamic behavior data; wherein, each tour path includes multiple consecutive nodes, and each node includes spatial coordinates, line - of - sight direction, and a preset stay duration;

[0010] S2: According to the line - of - sight directions of the nodes in the tour path, generate a continuous video sequence through the dynamic field - of - view chain algorithm;

[0011] S3: Based on the three - dimensional spatial form of the tour path, construct an environmental acoustic model through the ray - tracing method; according to the environmental acoustic model, simulate the sound at the nodes to generate a sound pressure level attenuation curve; wherein, the sound includes water sound, the sound of wind blowing through leaves, and the sound of birdsong;

[0012] S4: Generate a visual rhythm curve by extracting the time intervals of video frame switches from the video sequence; generate an acoustic scene rhythm curve by extracting the sound pressure level peak time points of the sound pressure level decay curve from the environmental acoustic model;

[0013] S5: Calculate the matching degree between the visual rhythm curve and the acoustic scene rhythm curve through the dynamic time warping algorithm. If the matching degree is lower than the first threshold, it is determined that the audiovisual coordination fails;

[0014] S6: If the audiovisual coordination fails, adjust the line-of-sight direction and the stay duration of the nodes in the tour path, and re-execute S2 to S5 until the matching degree is greater than or equal to the first threshold;

[0015] S7: Generate multi-sensory space sequence data according to the video sequence and the sound pressure level decay curve; wherein, the multi-sensory space sequence data includes standardized perception frames arranged in a spatio-temporal sequence, and each standardized perception frame contains visual perception data and auditory perception data corresponding to the nodes;

[0016] S8: Design the garden for the dynamic experience of tourists by mapping the multi-sensory space sequence data to a virtual reality interaction platform.

[0017] In a second aspect, the present application also provides a garden design system based on garden simulation, including:

[0018] A tourist path generation module, configured to generate multiple tour paths in the three-dimensional model of the garden based on the dynamic behavior data of tourists; wherein, each tour path includes multiple consecutive nodes, and each node includes spatial coordinates, a line-of-sight direction, and a preset stay duration;

[0019] A video sequence generation module, configured to receive the tour path and generate a continuous video sequence through the dynamic field-of-view chain algorithm according to the line-of-sight direction of the nodes in the tour path;

[0020] An acoustic modeling module, configured to receive the tour path; construct an environmental acoustic model through the ray tracing method based on the three-dimensional spatial form of the tour path; simulate the sound at the nodes according to the environmental acoustic model to generate a sound pressure level decay curve; wherein, the sound includes the sound of water, the sound of wind blowing through the leaves, and the sound of birdsong;

[0021] A rhythm curve generation module, configured to receive the video sequence and the sound pressure level decay curve; generate a visual rhythm curve by extracting the time intervals of video frame switches from the video sequence; generate an acoustic scene rhythm curve by extracting the sound pressure level peak time points of the sound pressure level decay curve from the environmental acoustic model;

[0022] An audiovisual collaboration analysis module, configured to receive a visual rhythm curve and a soundscape rhythm curve; calculate the matching degree between the visual rhythm curve and the soundscape rhythm curve through a dynamic time warping algorithm, and if the matching degree is lower than a first threshold, determine that the audiovisual collaboration fails and generate an optimization instruction;

[0023] A path optimization module, configured to respond to the optimization instruction, adjust the line-of-sight direction and the stay duration of nodes in the tour path, and sequentially mobilize a video sequence generation module, an acoustic modeling module, and a rhythm curve generation module to re-execute operations until the matching degree is greater than or equal to the first threshold;

[0024] A multi-sensory data synthesis module, configured to receive a video sequence and a sound pressure level attenuation curve, and generate multi-sensory spatial sequence data according to the video sequence and the sound pressure level attenuation curve; wherein, the multi-sensory spatial sequence data includes standardized perception frames arranged in a spatio-temporal sequence, and each standardized perception frame includes visual perception data and auditory perception data corresponding to the nodes;

[0025] A virtual reality mapping module, configured to receive the multi-sensory spatial sequence data, and design the garden for the dynamic experience of tourists by mapping the multi-sensory spatial sequence data to a virtual reality interaction platform.

[0026] In a third aspect, the present application further provides a computer device, including a memory and a processor, where the memory stores a computer program, and when the processor executes the computer program, it implements a garden design method based on garden simulation as in the first aspect.

[0027] In a fourth aspect, the present application further provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, it implements a garden design method based on garden simulation as in the first aspect.

[0028] The above-mentioned garden design method based on garden simulation generates a tour path including spatial coordinates, line-of-sight direction and stay duration based on tourists' dynamic behavior data, combines a dynamic field-of-view chain algorithm to generate a continuous video sequence, uses a ray tracing method to construct an environmental acoustic model to simulate a sound pressure level attenuation curve, extracts a visual rhythm curve and a soundscape rhythm curve and quantifies the matching degree by using a dynamic time warping algorithm, automatically iteratively adjusts path parameters until the audiovisual collaboration meets the standard, and finally generates spatio-temporally aligned multi-sensory spatial sequence data and maps it to a virtual reality platform, solving the problems of fragmented audiovisual simulation, low accuracy of acoustic models, and poor efficiency relying on manual adjustment in traditional garden design, realizing high-precision collaborative simulation and automated design optimization of multi-sensory experience in a dynamic tour scenario, and significantly improving the authenticity of the immersive experience and the efficiency of scheme generation. Brief Description of the Drawings

[0029] To more clearly illustrate the technical solutions in the embodiments of the present application or related technologies, the following will briefly introduce the drawings required for the description of the embodiments or related technologies. Obviously, the drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0030] Figure 1 Schematic flow chart of a garden design method based on garden simulation provided by the present invention;

[0031] Figure 2 Schematic structural diagram of a garden design system based on garden simulation provided by the present invention. Detailed implementation manners

[0032] In order to make the objectives, technical solutions and advantages of the present application more clear, the following further details the present application in conjunction with the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0033] Refer to Figure 1 , which shows a schematic flow chart of a garden design method based on garden simulation provided by the present application. The method includes the following steps:

[0034] S1: In the three-dimensional model of the garden, generate multiple tour paths based on the dynamic behavior data of tourists; wherein, each tour path includes multiple consecutive nodes, and each node includes spatial coordinates, line-of-sight direction and a preset stay duration.

[0035] Specifically, first use professional three-dimensional modeling software (such as 3ds Max, SketchUp, etc.) to accurately construct a three-dimensional model of the garden according to the actual terrain, buildings, vegetation and other elements of the garden. This model should include all important landscape elements in the garden, such as rockeries, pools, pavilions, paths, etc., and the parameters such as the position and size of each element should be consistent with the actual situation, providing an accurate basic scene for subsequent simulation.

[0036] Collect the dynamic behavior data of tourists in the garden through various methods. For example, conduct on-site observations and records in a real garden, and count information such as the walking routes, stay times, and line-of-sight directions of tourists in different areas; questionnaires or interviews can also be used to understand tourists' preferences for garden landscapes and tour habits. In addition, the tourist behavior research data of similar gardens in the past can also be referred to as supplements and references.

[0037] Based on the collected dynamic behavior data of tourists, multiple possible tour paths are generated in the 3D model using specific path generation algorithms (such as ant colony algorithm, genetic algorithm, etc.). These algorithms comprehensively consider factors such as tourists' behavior preferences and the attractiveness of landscapes, making the generated tour paths more in line with the actual tour needs of tourists. Each tour path consists of multiple consecutive nodes, and the density of the nodes can be set according to the actual situation to describe in detail the movement trajectory of tourists in the garden. Each node contains spatial coordinates (x, y, z) to determine the specific position of tourists in the 3D model; the line-of-sight direction (usually represented by angles in the horizontal and vertical directions), reflecting the focus of attention and viewing direction of tourists at this position; and a preset stay duration (estimated based on the average stay time of tourists in front of similar landscapes), used to simulate the viewing and experience time of tourists for the surrounding landscapes at this node.

[0038] S2: According to the line-of-sight directions of the nodes in the tour path, a continuous video sequence is generated through the dynamic field-of-view chain algorithm.

[0039] Specifically, the dynamic field-of-view chain algorithm is an algorithm that can dynamically calculate the landscape range and changes that tourists can view at different positions based on the line-of-sight directions and movement trajectories of tourists. It regards the line of sight of tourists as a ray, starting from the position of the tourists' eyes (node spatial coordinates), extending along the line-of-sight direction, and intersecting with various landscape elements in the 3D model, thereby determining the landscape content and range that tourists can see at this node.

[0040] According to the tour path generated in S1, starting from the starting node, in the order and stay duration of each node, the dynamic field-of-view chain algorithm is used to calculate the visible area of tourists at each node in turn. At each node, corresponding image frames are generated based on the landscape elements in the visible area and their relative positions, shapes, colors, etc. Then, these image frames are arranged in chronological order to form a continuous video sequence, which can simulate the visual experience of tourists during the tour, as if tourists were on the spot in the garden, enjoying the scenery along the tour path.

[0041] S3: Based on the three-dimensional spatial form of the tour path, an environmental acoustics model is constructed through the ray tracing method; according to the environmental acoustics model, the sound at the nodes is simulated to generate a sound pressure level attenuation curve; where the sounds include water sounds, the sound of wind blowing through leaves, and bird calls.

[0042] Specifically, ray tracing is a method commonly used in acoustic simulation, which simulates the propagation of sound in space as the propagation of rays. In the three-dimensional model of the garden, according to the three-dimensional spatial form of the tour path, a large number of rays are emitted from each node in all directions to simulate the propagation path of sound in the garden. During the propagation process, these rays will reflect, refract, absorb and other acoustic phenomena with the buildings, vegetation, terrain and other elements in the garden. By calculating the propagation characteristics of these rays, an acoustic model of the entire garden environment can be constructed.

[0043] Considering the common natural sounds in the garden, the sound of water, wind blowing leaves and birds singing are selected as the main sound sources. The audio materials of these sound sources can be obtained through on-site recording or from professional audio libraries. For each sound source, it is necessary to perform corresponding audio processing according to its actual position and propagation characteristics in the garden, such as adjusting the volume, pitch, reverberation and other parameters to make it more in line with the actual acoustic environment of the garden.

[0044] Based on the constructed environmental acoustic model, the sound at each node is simulated and calculated. By analyzing the changes in the sound pressure level of the sound at different distances and under different obstacles, the sound pressure level attenuation curve is generated. These curves can intuitively reflect the propagation and attenuation laws of sound in the garden space, providing data support for the subsequent generation of soundscape rhythm curves.

[0045] S4: Generate a visual rhythm curve by extracting the time interval of video frame switching from the video sequence; generate a soundscape rhythm curve by extracting the sound pressure level peak time point of the sound pressure level attenuation curve from the environmental acoustic model.

[0046] Specifically, the time intervals of video frame switching are extracted from the video sequence generated by S2. The length of these time intervals reflects the changes in the visual rhythm of the video content. For example, when the time interval is short, the screen switches frequently and the visual rhythm is brisk; on the contrary, when the time interval is long, the screen is relatively stable and the visual rhythm is slow. Based on the data of these time intervals, a visual rhythm curve is drawn. The horizontal axis of the curve represents time, and the vertical axis represents the frequency or rhythm intensity of video frame switching. The changes in the visual rhythm during the tour can be intuitively seen through the ups and downs of the curve.

[0047] From the sound pressure level attenuation curve generated in S3, extract the time points of the peak sound pressure level. These peak time points usually correspond to the sudden changes or emphasized parts of the sound, such as the splashing of water and the high-pitched part of the bird's song, which form a distinct sense of rhythm in hearing. Arrange these sound pressure level peak time points in chronological order to generate a soundscape rhythm curve. The horizontal axis of the curve is time, and the vertical axis is the intensity of the sound pressure level peak. The fluctuation of the curve can show the rhythm changes of the soundscape during the tour.

[0048] S5: Calculate the matching degree between the visual rhythm curve and the soundscape rhythm curve through the dynamic time warping algorithm. If the matching degree is lower than the first threshold, it is determined that the audiovisual coordination fails.

[0049] Specifically, the dynamic time warping algorithm (DTW) is an algorithm for comparing the similarity of two time series. It can stretch or compress the time axis to make the two time series achieve the best match in shape. In this method, it is used to calculate the matching degree between the visual rhythm curve and the soundscape rhythm curve to evaluate the coordination between vision and hearing during the tour.

[0050] Take the visual rhythm curve and the soundscape rhythm curve generated in S4 as two time series and input them into the DTW algorithm. The algorithm will calculate the matching degree value between them. The size of this matching degree value reflects the similarity degree of the two curves. The higher the matching degree, the better the coordination between vision and soundscape in terms of rhythm; on the contrary, a low matching degree indicates poor coordination. Set a first threshold. When the calculated matching degree is lower than this threshold, it is determined that the audiovisual coordination fails, that is, it is considered that the visual and auditory experiences on this tour path in the garden design do not achieve a good coordination effect and need to be adjusted and optimized.

[0051] S6: If the audiovisual coordination fails, adjust the line-of-sight direction and the staying duration of the nodes in the tour path, and re-execute S2 to S5 until the matching degree is greater than or equal to the first threshold.

[0052] Specifically, when it is determined that the audiovisual coordination fails, the tour path needs to be adjusted. The main parameters for adjustment are the line-of-sight direction and the staying duration of the nodes. For the adjustment of the line-of-sight direction, the relationship between the position of the sound source and the tourist's line-of-sight direction can be analyzed according to the time point of the sound pressure level peak in the soundscape rhythm curve, and the line-of-sight direction can be appropriately adjusted so that the tourist can see the landscape elements corresponding to the sound at the appropriate time, enhancing the audiovisual coordination effect. For the adjustment of the staying duration, according to the matching situation between the visual rhythm curve and the soundscape rhythm curve, at the nodes where the visual and auditory rhythms change significantly, the staying time can be appropriately extended or shortened so that the tourist has enough time to feel and appreciate the integration of the landscape and the sound.

[0053] After adjusting the tour path parameters, re-execute the steps of S2 to S5, that is, regenerate the video sequence, construct the environmental acoustic model, simulate the sound, generate the visual and soundscape rhythm curves, and calculate the matching degree. Through multiple iterative adjustments and simulations, until the matching degree is greater than or equal to the first threshold, ensure that the audiovisual coordination reaches a good effect.

[0054] S7: Generate multi-sensory spatial sequence data based on the video sequence and the sound pressure level decay curve; wherein, the multi-sensory spatial sequence data includes standardized perception frames arranged in a spatio-temporal sequence, and each standardized perception frame contains visual perception data and auditory perception data corresponding to the nodes.

[0055] Specifically, the multi-sensory spatial sequence data is a data form that integrates visual and auditory perception data in a spatio-temporal sequence. It not only contains the visual information of tourists at each node on the tour path (such as the landscape pictures in the video sequence), but also contains the corresponding auditory information (such as the sound data in the sound pressure level decay curve). Through this data form, the changes in the multi-sensory experience of tourists during the garden tour can be described comprehensively and systematically, providing richer data support for garden design.

[0056] Extract the corresponding data from the video sequence in S2 and the sound pressure level decay curve in S3 to construct standardized perception frames. Each perception frame corresponds to a node and contains the visual perception data of the node (such as characteristic parameters of the picture color, brightness, composition, etc.) and the auditory perception data (such as characteristic parameters of the sound frequency, sound pressure level, timbre, etc.). Standardize these perception data so that they can be compared and analyzed on the same scale and within the same range, in order to better integrate and utilize these multi-sensory data.

[0057] S8: Design the garden according to the dynamic experience of tourists by mapping the multi-sensory spatial sequence data to a virtual reality interaction platform.

[0058] Specifically, select a suitable virtual reality interaction platform (such as Unity, Unreal Engine, etc.), and map the multi-sensory spatial sequence data generated in S7 to this platform. Build a virtual garden scene on the platform, enabling designers to experience the tour process of tourists in the garden in an immersive manner, including multi-sensory experiences such as the landscapes seen and the sounds heard.

[0059] Designers design and optimize the garden on the virtual reality interaction platform according to the dynamic experience of tourists. For example, according to the situation shown in the multi-sensory spatial sequence data that tourists stay longer at certain nodes and have better audio-visual synergy effects, the landscape features of this area can be further strengthened; for areas with ineffective audio-visual synergy or poor tourist experience, the landscape layout can be adjusted, some elements can be added or reduced, the position or type of the sound source can be changed, etc. Through continuous iterative design and simulation, a garden design plan that can provide tourists with a good multi-sensory experience can be finally obtained. At the same time, the interactive function of the virtual reality platform can also be used to let tourists experience the garden design effect in advance, collect the feedback opinions of tourists, and further optimize the design plan.

[0060] The above-mentioned garden design method based on garden simulation generates a tour path including spatial coordinates, line-of-sight direction, and stay duration based on tourists' dynamic behavior data, combines the dynamic visual field chain algorithm to generate a continuous video sequence, uses the ray tracing method to construct an environmental acoustic model to simulate the sound pressure level attenuation curve, extracts the visual rhythm curve and the soundscape rhythm curve, and uses the dynamic time warping algorithm to quantify the matching degree, automatically iteratively adjusts the path parameters until the audio-visual coordination meets the standard, and finally generates multi-sensory spatial sequence data with spatio-temporal alignment and maps it to a virtual reality platform, solving the problems of fragmented audio-visual simulation, low accuracy of acoustic models, and poor efficiency relying on manual adjustment in traditional garden design, realizing high-precision coordination simulation and automated design optimization of multi-sensory experiences in dynamic tour scenarios, and significantly improving the authenticity of immersive experiences and the efficiency of scheme generation.

[0061] In an alternative embodiment, S2 includes the following steps:

[0062] S21: Obtain the line-of-sight direction data of the current node according to the line-of-sight direction of the nodes in the tour path.

[0063] Specifically, in a three-dimensional model, the line-of-sight direction can be represented by the horizontal angle (azimuth) and the vertical angle (elevation). The horizontal angle refers to the angle formed by rotating clockwise from the due north direction to the projection of the line of sight on the horizontal plane, and the range is usually 0° to 360°; the vertical angle refers to the angle between the line of sight and the horizontal plane, and the range is generally -90° to 90°. For example, when a tourist looks straight ahead at the due north, the horizontal angle is 0° and the vertical angle is 0°; when a tourist looks up at the sky in the due north, the horizontal angle is still 0°, and the vertical angle is positive; when a tourist looks down at the ground in the due north, the horizontal angle is 0°, and the vertical angle is negative.

[0064] These line-of-sight direction data are obtained based on the previous observation and analysis of tourists' dynamic behavior. In an actual garden scene, by observing the head orientation and line-of-sight focus of tourists when viewing landscapes at different positions, the line-of-sight direction of each node can be roughly determined. In addition, professional equipment such as an eye tracker can also be used to test tourists in an actual garden or virtual reality environment to accurately record the line-of-sight direction data of tourists at each node.

[0065] S22: Based on the line-of-sight direction data, obtain the visual field range of the current node through the dynamic visual field chain algorithm, and calculate the visible area boundary of the visual field range.

[0066] Specifically, the dynamic field of view chain algorithm, based on the line-of-sight direction data of the current node, emits a series of rays in the 3D model starting from the position of this node along the line-of-sight direction to simulate the range that the tourist's line of sight can reach. These rays will intersect with various elements in the garden (such as buildings, trees, rockeries, etc.) during the propagation process, and the algorithm will record the information of these intersection points. Through these intersection points, the field of view range of the tourist at this node can be determined, that is, the landscape area that the tourist can directly see.

[0067] Based on all the intersection points within the field of view range, geometric calculation methods (such as the convex hull algorithm or the contour line extraction algorithm) are used to calculate the boundary of the visible area. The boundary of the visible area is a closed curve that separates the area within the field of view range from the area outside the field of view range. This boundary can be represented by a series of continuous points or line segments in the 3D model, providing an accurate visual range limit for the subsequent generation of video sequences.

[0068] S23: Generate a continuous video sequence frame by frame according to the boundary of the visible area and the preset stay duration of the node.

[0069] Specifically, the frame rate is determined according to the preset stay duration of the node and the desired smoothness of the video. For example, if the preset stay duration is 10 seconds and the desired frame rate is 30 frames per second, then 300 frames of images need to be generated within these 10 seconds. According to the total number of frames and the stay duration, the time interval between each frame is calculated, that is, 1 / 30 second. This time interval determines the speed of the picture switching in the video sequence.

[0070] At each time interval, according to the line-of-sight direction and the field of view of the current node, the landscape elements within the visible area at that moment are extracted from the 3D model, and corresponding image frames are generated according to certain rendering rules (such as lighting, color, texture, etc.). The rendered image frames should have a visual effect similar to the actual garden scene, including details such as the shape, size, color, and brightness of the landscape, so that the generated video sequence can truly reflect the tourist's visual experience at this node. Then, these image frames are arranged in chronological order to form a continuous video clip, and the video clips of all nodes are connected in sequence to obtain the continuous video sequence of the entire tour path.

[0071] S24: Calculate the overlap rate of the field of view ranges of adjacent nodes. If the overlap rate is lower than the second threshold, insert a transition viewpoint within the field of view range and regenerate the video sequence.

[0072] Specifically, for two adjacent nodes, their field of view ranges are obtained respectively. Calculate the ratio of the area of the overlapping region of these two field of view ranges to the area of the field of view range of the previous node, which is the overlap rate. For example, if the area of the field of view range of the previous node is S1 and the area of the overlapping region is S2, then the overlap rate = S2 / S1. The overlap rate reflects the degree of visual continuity between adjacent nodes. The higher the overlap rate, the more natural the field of view connection between the two nodes, and the smoother the transition of the visual experience of tourists during movement.

[0073] If the overlap rate is lower than the second threshold, it means that the field of view connection between adjacent nodes is not natural enough, which may lead to obvious jumps or discontinuities in the video sequence. In this case, transitional viewpoints need to be inserted within the field of view range between the two nodes. The number and positions of the transitional viewpoints can be determined according to the changes in the field of view and the desired transitional effect. After inserting the transitional viewpoints, recalculate the field of view ranges of these transitional viewpoints, and regenerate the video sequence according to the new field of view ranges and time intervals. In this way, it can be ensured that the transition of the video sequence between adjacent nodes is smoother, improving the coherence and comfort of the tourists' visual experience.

[0074] In an alternative embodiment, S3 includes the following steps:

[0075] S31: Based on the three-dimensional spatial form of the tour path, construct the geometric topology of the environmental acoustic model, and extract the reflection surface parameters of the geometric topology; wherein, the reflection surface parameters include surface curvature and material type.

[0076] Specifically, based on the three-dimensional spatial form of the tour path, use three-dimensional modeling software or a custom modeling algorithm to construct the geometric topology of the environmental acoustic model. This geometric topology accurately describes the positions, shapes, and spatial relationships between various interfaces (such as walls, floors, ceilings, vegetation surfaces, etc.) in the garden. For example, for a garden with complex terrain and buildings, through precise modeling, the three-dimensional coordinate information of the interfaces such as the walls and roofs of each building, as well as the surface geometry formed by the undulations of the terrain, can be obtained.

[0077] In the constructed geometric topology, identify and extract all reflection surface parameters that may affect sound reflection. The reflection surface parameters mainly include surface curvature and material type. Surface curvature reflects the degree of curvature of the reflection surface, which affects the direction and energy distribution of sound during reflection. For example, a convex surface will cause the sound to spread in all directions after reflection, while a concave surface may cause the sound to be focused and reflected to a certain area. The material type determines the acoustic characteristics of the reflection surface. Different materials have different absorption and reflection abilities for sound. For example, materials such as brick walls, stones, woods, and glasses have different sound absorption and reflection characteristics.

[0078] S32: According to a preset material acoustic parameter library, map the reflector parameters to the sound absorption coefficient and scattering rate.

[0079] Specifically, establish a preset material acoustic parameter library, which stores the sound absorption coefficient and scattering rate data of various common materials at different frequencies. These data can be obtained through experimental measurements, referring to acoustic material manuals, or from professional acoustic databases. For example, for a brick wall material, the sound absorption coefficient may be relatively low in the low-frequency band, while it will increase in the high-frequency band; the sound absorption coefficient of wood is relatively high, especially in the mid-high frequency band, with good sound absorption performance.

[0080] According to the extracted material type of the reflector, find the corresponding sound absorption coefficient and scattering rate from the material acoustic parameter library, and map these values to the corresponding reflector. For example, if the material of a reflector is stone, then obtain the sound absorption coefficient and scattering rate of stone at each frequency from the parameter library and assign them to this reflector. This mapping process converts the physical characteristics of the reflector into the parameters required for acoustic simulation, providing a basis for subsequent acoustic wave propagation simulation.

[0081] S33: By using the ray tracing method, simulate the propagation path of acoustic waves in the environmental acoustic model based on the sound absorption coefficient and scattering rate, and count the number of reflections and attenuation duration of the propagation path.

[0082] Specifically, based on the ray tracing method, take each node as the sound source position and emit a large number of rays in all directions to simulate the propagation of acoustic waves. These rays will undergo reflection, refraction, absorption and other phenomena when propagating in the environmental acoustic model. By calculating parameters such as the reflection angle, energy attenuation and propagation time of the rays on each reflector, trace the propagation path of the rays until the energy of the rays decays below a certain set threshold or reaches the maximum propagation number. For example, when a ray is emitted from a node and hits the first reflector, calculate the reflection direction and energy attenuation of the ray according to the law of reflection and the sound absorption coefficient and scattering rate of the material, and then continue to trace the reflected ray, recording its subsequent propagation path, number of reflections and total attenuation duration.

[0083] During the ray tracing process, for each ray, count the number of reflections and total attenuation duration experienced from the sound source until the energy decays below the threshold. The number of reflections reflects the frequency of interaction between acoustic waves and reflectors when propagating in the environment, and the attenuation duration represents the time required for the acoustic wave energy to decay from the initial value to a certain level. Through statistical analysis of a large number of rays, statistical parameters such as the average number of reflections and average attenuation duration of acoustic wave propagation in the entire environmental acoustic model can be obtained. These parameters can reflect the propagation characteristics and energy attenuation law of acoustic waves in the garden space.

[0084] S34: Generate a sound pressure level attenuation curve based on the number of reflections and the decay duration.

[0085] Specifically, based on the statistically obtained number of reflections and decay duration data, combined with the energy attenuation formula of sound waves during propagation, calculate the sound pressure level at different time points or different propagation distances. Generally, the sound pressure level is proportional to the energy of the sound wave. As the number of reflections increases and the decay duration prolongs, the sound pressure level will gradually decrease. Taking time as the horizontal axis and the sound pressure level as the vertical axis, connect the calculated data points to form a sound pressure level attenuation curve. This curve intuitively shows the change trend of the sound pressure level over time or the propagation path, reflecting the attenuation process of sound in the garden environment.

[0086] By analyzing the sound pressure level attenuation curve, the propagation effect of sound in the garden can be understood. For example, if the initial segment of the curve is relatively steep, it indicates that at the initial stage of sound propagation, the energy decays rapidly, which may be due to the strong absorption effect of the reflecting surface or a large number of reflections; the flat segment of the curve indicates that after multiple reflections and attenuations, the sound tends to a stable state. Designers can evaluate the sound environment quality of different areas according to the sound pressure level attenuation curve. For example, near a water feature, it is desired that the sound of water can be clearly heard within a certain range but not be too noisy. By adjusting the material or layout of the reflecting surface, the sound pressure level attenuation curve can be made to reach an ideal state, thereby optimizing the soundscape design of the garden.

[0087] In an optional embodiment, S5 includes the following steps:

[0088] S51: Extract the video frame switching timestamp sequence from the visual rhythm curve and extract the sound pressure level peak timestamp sequence from the soundscape rhythm curve.

[0089] Specifically, from the visual rhythm curve, extract the timestamp sequence of video frame switching. Each timestamp in this sequence represents the moment of video frame switching, which reflects the update frequency and rhythm change of visual information. For example, in a video sequence, if video frame switching occurs at 0 seconds, 0.5 seconds, 1 second, 1.5 seconds, etc., then the time values at these moments form the video frame switching timestamp sequence. These timestamp sequences can intuitively show the speed and change of the visual rhythm, providing visual time reference points for subsequent matching degree calculation.

[0090] From the soundscape rhythm curve, extract the timestamp sequence of the sound pressure level peak. These timestamps correspond to the peak moments of sound intensity, usually important events or feature points in the sound, such as the moment when water splashes or the high-pitched part of a bird's song. By extracting these sound pressure level peak timestamps, the key time nodes of the sound rhythm can be obtained for comparison and matching analysis with the visual rhythm.

[0091] S52: Normalize the video frame switching timestamp sequence and the sound pressure level peak timestamp sequence to generate a normalized timestamp sequence.

[0092] Specifically, to eliminate the differences in dimension and magnitude between different timestamp sequences and enable the timestamp sequences of vision and soundscape to be compared and analyzed on the same scale, it is necessary to normalize the extracted video frame switching timestamp sequence and the sound pressure level peak timestamp sequence. Common normalization methods include min-max normalization, mean-standard deviation normalization, etc. For example, using the min-max normalization method, map the time values in each timestamp sequence to the interval [0, 1], where 0 represents the start time of the sequence and 1 represents the end time of the sequence.

[0093] After normalization, a normalized timestamp sequence for video frame switching and a normalized timestamp sequence for sound pressure level peak are generated respectively. These two normalized timestamp sequences convert time information into relative proportional values, facilitating the subsequent dynamic time warping algorithm to align and compare the two time series to evaluate the matching degree of vision and soundscape in rhythm.

[0094] S53: According to the normalized timestamp sequence, calculate the alignment path distance through the dynamic time warping algorithm to obtain the minimum alignment path distance, and use the minimum alignment path distance as a quantitative index of the matching degree.

[0095] Specifically, the dynamic time warping algorithm (DTW) is an algorithm used to measure the similarity between two time series. It finds the optimal alignment path between two time series to make the points on the path correspond to similar features in time as much as possible. In this method, the normalized video frame switching timestamp sequence and the sound pressure level peak timestamp sequence are used as the two time series and input into the DTW algorithm. The algorithm constructs an accumulated distance matrix, and each element in the matrix represents the accumulated distance when aligning the i-th point in the video frame switching timestamp sequence with the j-th point in the sound pressure level peak timestamp sequence. Then, through the method of dynamic programming, find a path from the starting point (0, 0) to the ending point (n, m) (n and m are the lengths of the two time series respectively) in the matrix, so that the sum of the accumulated distances on the path is the smallest. This path is the optimal alignment path, and the corresponding sum of the accumulated distances is the minimum alignment path distance.

[0096] The minimum alignment path distance reflects the matching degree of the two time series in rhythm. The smaller the distance, the more similar the rhythms of the two time series, and the better the coordination between vision and soundscape; conversely, the larger the distance, the greater the rhythm difference and the worse the coordination. Therefore, using the minimum alignment path distance as a quantitative index of the matching degree can objectively evaluate the effect of audiovisual coordination and provide a quantitative basis for subsequent judgment and adjustment.

[0097] S54: If the value of the minimum alignment path distance exceeds a preset distance threshold, it is determined that the audio-visual collaboration fails; wherein, the preset distance threshold is the first threshold.

[0098] Specifically, the setting of the preset distance threshold (i.e., the first threshold) is determined according to the actual garden design requirements and tourist experience standards. This threshold represents the acceptable matching degree of rhythm between vision and soundscape. When the minimum alignment path distance exceeds this threshold, it means that the rhythm difference between vision and soundscape exceeds the acceptable range, which may cause tourists to feel the disharmony between vision and hearing during the tour, thus affecting the overall tour experience.

[0099] If it is determined that the audio-visual collaboration fails, it is necessary to adjust and optimize the tour path. The adjustment method can refer to the steps in S6. By modifying parameters such as the line-of-sight direction and stay duration of the nodes, regenerating the video sequence and the soundscape rhythm curve, and calculating the matching degree again until the minimum alignment path distance is less than or equal to the preset distance threshold, ensuring a good collaborative effect between vision and soundscape in terms of rhythm and providing tourists with a comfortable and natural multi-sensory experience.

[0100] In an alternative embodiment, S7 includes the following steps:

[0101] S71: Perform a frustum analysis on the video sequence, extract the RGB-D data and texture features of the landscape elements within the viewing angle of each node, and generate visual perception data containing spatial depth information.

[0102] Specifically, frustum analysis is a method for determining the position and distribution of landscape elements within the visible area of each node in the video sequence. By processing the image frames of each node in the video sequence, according to the internal parameters (such as focal length, image sensor size, etc.) and external parameters (such as position, line-of-sight direction, etc.) of the camera, calculate the opening angle and spatial range of the frustum. Within this spatial range, identify and extract each landscape element, such as trees, flowers, buildings, rockeries, etc., and record their three-dimensional coordinate positions.

[0103] RGB-D data refers to data containing red (R), green (G), blue (B) color information and depth (D) information. Through a depth sensor or computer vision algorithm, extract the RGB-D data of the landscape elements within the viewing angle of each node from the video sequence. The depth information can provide the spatial distance between the landscape elements and the tourists, enhancing the three-dimensional sense and spatial sense of the visual perception data. At the same time, extract the texture features of the landscape elements, such as the bark texture of trees, the petal texture of flowers, etc. These texture features can enrich the detailed information of the visual perception data, making the generated visual perception data more real and vivid.

[0104] Integrate the extracted RGB-D data and texture features to generate visual perception data containing spatial depth information. The spatial depth information reflects the distance relationship of landscape elements through the magnitude of depth values, enabling designers to clearly understand the spatial layout of landscape elements that tourists can see at each node. For example, at a node, tourists may see nearby flowers (with smaller depth values) and distant buildings (with larger depth values), and all this information is included in the visual perception data, providing detailed visual information for subsequent generation of multi-sensory spatial sequence data and garden design.

[0105] S72: Classify and identify the reverberation components of underwater sound, the sound of wind blowing through leaves, and bird calls according to the spectral characteristics of the sound pressure level decay curve, and generate auditory perception data containing the sound source type and azimuth vector.

[0106] Specifically, perform spectral analysis on the sound pressure level decay curve to decompose the sound signal into amplitude and phase information of different frequency components. According to the spectral characteristic differences among underwater sound, the sound of wind blowing through leaves, and bird calls, use signal processing algorithms (such as filtering, feature extraction, etc.) to classify and identify these sounds. For example, underwater sound usually has lower frequency components and a relatively continuous spectrum; the frequency components of the sound of wind blowing through leaves are relatively higher and have a certain degree of randomness; bird calls have unique frequency modulation and amplitude variation characteristics.

[0107] While classifying and identifying different types of sounds, analyze the reverberation components in the sounds. Reverberation refers to the complex sound phenomenon formed after sound waves are reflected multiple times in space, which can provide additional information about spatial characteristics (such as space size, shape, material, etc.). Through acoustic models and signal processing techniques, separate the reverberation components in underwater sound, the sound of wind blowing through leaves, and bird calls, and quantitatively analyze parameters such as reverberation time and reverberation intensity. The analysis results of these reverberation components help to more accurately describe the propagation characteristics and sense of space of sound in the garden environment.

[0108] Integrate the classified and identified underwater sound, the sound of wind blowing through leaves, and bird calls, as well as the analysis results of their reverberation components, to generate auditory perception data containing the sound source type and azimuth vector. The azimuth vector represents the direction and position information of the sound, which is calculated through sound source localization algorithms (such as based on time difference of arrival, phase difference, etc.) of sound waves. For example, at a node, tourists may hear underwater sound on the left (azimuth vector pointing to the left), bird calls in the front (azimuth vector pointing to the front), and the sound of wind blowing through leaves around (azimuth vectors scattered in all directions), and all this information is included in the auditory perception data, providing a rich auditory experience description for tourists.

[0109] S73: Align the visual perception data and auditory perception data according to the time and space stamps to generate a standardized perception frame.

[0110] Specifically, both visual perception data and auditory perception data are data sequences with timestamps. By matching the timestamps, the two are aligned on the time axis. At the same time, considering the spatial movement of tourists during the tour, according to the spatial position information of the nodes, the visual and auditory data are also aligned spatially. For example, at a specific time point, the tourist is located at a certain node, and the visual perception data and auditory perception data at this time both correspond to the spatial position of this node. Through spatio-temporal stamp alignment, the consistency of visual and auditory data in time and space is ensured, so that the generated standardized perception frames can accurately reflect the multi-sensory experience of tourists at each node.

[0111] Integrate the aligned visual perception data and auditory perception data into a data structure to form standardized perception frames. Each standardized perception frame contains all visual and auditory perception information corresponding to the nodes, such as RGB-D data of landscape elements, texture features, sound source types, azimuth vectors, etc. This standardized data format facilitates subsequent data processing and analysis, and also enables multi-sensory spatial sequence data to have a unified interface and specification, which can be better applied to virtual reality interaction platforms and other garden design tools. For example, on a virtual reality platform, by reading the data in the standardized perception frames, the landscape pictures seen and the sounds heard by tourists at each node can be presented simultaneously, providing designers with an intuitive and comprehensive tool for multi-sensory experience evaluation and design optimization.

[0112] The above-mentioned garden design method based on garden simulation achieves the technical effects of optimizing garden design and enhancing tourists' multi-sensory experience by integrating various technical means such as tourists' dynamic behavior data, 3D modeling, dynamic field-of-view chain algorithm, ray tracing method, dynamic time warping algorithm, and virtual reality interaction platform. Specifically, use tourists' dynamic behavior data to generate a tour path containing spatial coordinates, line-of-sight directions, and preset stay durations, combine the dynamic field-of-view chain algorithm to simulate tourists' visual experience and generate a video sequence, and use the ray tracing method to construct an environmental acoustic model to simulate sound propagation and generate a sound pressure level attenuation curve. Then, through the dynamic time warping algorithm, evaluate the matching degree between the visual rhythm curve and the soundscape rhythm curve, realize the determination and optimization adjustment of the audiovisual coordination effect, and finally map the multi-sensory spatial sequence data to the virtual reality interaction platform to provide tourists with an immersive garden design experience, making the garden design more in line with tourists' dynamic perception needs and enhancing the overall tour effect.

[0113] It should be understood that although the steps in the flowcharts involved in the above-described embodiments are sequentially shown according to the indications of the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless specifically stated herein, there is no strict order restriction for the execution of these steps, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above-described embodiments may include multiple steps or multiple stages. These steps or stages are not necessarily executed at the same moment, but can be executed at different moments, and the execution order of these steps or stages is not necessarily sequential, but can be executed alternately or in turn with at least a part of other steps or steps or stages in other steps.

[0114] Based on the same inventive concept, an embodiment of the present application further provides a system for implementing the above-mentioned garden design method based on garden simulation. The implementation solutions for solving problems provided by this system are similar to the implementation solutions described in the above method. Therefore, the specific limitations in one or more embodiments of the following garden design system based on garden simulation can refer to the limitations on a garden design method based on garden simulation in the foregoing, and will not be repeated here.

[0115] In an exemplary embodiment, as Figure 2 shown, a garden design system 20 based on garden simulation is provided, including:

[0116] A tourist path generation module 21, configured to generate multiple tour paths in the three-dimensional model of the garden based on tourist dynamic behavior data; wherein each tour path includes multiple consecutive nodes, and each node includes spatial coordinates, a line-of-sight direction, and a preset stay duration.

[0117] A video sequence generation module 22, configured to receive the tour path and generate a continuous video sequence through a dynamic field-of-view chain algorithm according to the line-of-sight direction of the nodes in the tour path.

[0118] An acoustic modeling module 23, configured to receive the tour path; construct an environmental acoustic model through a ray tracing method based on the three-dimensional spatial form of the tour path; simulate the sound at the nodes according to the environmental acoustic model to generate a sound pressure level attenuation curve; wherein the sound includes water sound, the sound of wind blowing through leaves, and the sound of birdsong.

[0119] A rhythm curve generation module 24, configured to receive the video sequence and the sound pressure level attenuation curve; generate a visual rhythm curve by extracting the time interval of video frame switching from the video sequence; generate a soundscape rhythm curve by extracting the sound pressure level peak time points of the sound pressure level attenuation curve from the environmental acoustic model.

[0120] The audiovisual collaborative analysis module 25 is used to receive the visual rhythm curve and the soundscape rhythm curve; calculate the matching degree between the visual rhythm curve and the soundscape rhythm curve through the dynamic time warping algorithm. If the matching degree is lower than the first threshold, it is determined that the audiovisual collaboration fails and an optimization instruction is generated.

[0121] The path optimization module 26 is used to respond to the optimization instruction, adjust the line of sight direction and the staying duration of the nodes in the tour path, and sequentially mobilize the video sequence generation module, the acoustic modeling module, and the rhythm curve generation module to re-execute the operations until the matching degree is greater than or equal to the first threshold.

[0122] The multi-sensory data synthesis module 27 is used to receive the video sequence and the sound pressure level attenuation curve, and generate multi-sensory space sequence data according to the video sequence and the sound pressure level attenuation curve; wherein, the multi-sensory space sequence data includes standardized perception frames arranged in a spatio-temporal sequence, and each standardized perception frame contains visual perception data and auditory perception data corresponding to the nodes.

[0123] The virtual reality mapping module 28 is used to receive the multi-sensory space sequence data, and design the garden for the dynamic experience of tourists by mapping the multi-sensory space sequence data to the virtual reality interaction platform.

[0124] Optionally, the video sequence generation module 22 includes:

[0125] The line of sight direction acquisition unit 221 is used to receive the tour path and obtain the line of sight direction data of the current node according to the line of sight direction of the nodes in the tour path.

[0126] The visible area calculation unit 222 is used to receive the line of sight direction data, and based on the line of sight direction data, obtain the field of view range of the current node through the dynamic field of view chain algorithm and calculate the visible area boundary of the field of view range.

[0127] The frame-by-frame rendering unit 223 is used to receive the visible area boundary and the preset staying duration of the node, and generate a continuous video sequence frame by frame according to the visible area boundary and the preset staying duration of the node.

[0128] The transition viewpoint insertion unit 224 is used to receive the visible area boundary, calculate the overlap rate of the field of view ranges of adjacent nodes. If the overlap rate is lower than the second threshold, insert a transition viewpoint in the field of view range and mobilize the frame-by-frame rendering unit to regenerate the video sequence.

[0129] Optionally, the acoustic modeling module 23 includes:

[0130] The geometric topology construction unit 231 is used to receive the tour path, construct the geometric topology structure of the environmental acoustic model based on the three-dimensional spatial form of the tour path, and extract the reflection surface parameters of the geometric topology structure; wherein, the reflection surface parameters include surface curvature and material type.

[0131] An acoustic parameter mapping unit 232, configured to receive the reflector parameters and map the reflector parameters to the sound absorption coefficient and the scattering rate according to a preset material acoustic parameter library.

[0132] An acoustic wave propagation simulation unit 233, configured to receive the sound absorption coefficient and the scattering rate; simulate the propagation path of the acoustic wave in the environmental acoustic model based on the sound absorption coefficient and the scattering rate by using the ray tracing method, and count the number of reflections and the attenuation duration of the propagation path.

[0133] A sound pressure level analysis unit 234, configured to receive the number of reflections and the attenuation duration and generate a sound pressure level attenuation curve according to the number of reflections and the attenuation duration.

[0134] Optionally, the audiovisual collaboration analysis module 25 includes:

[0135] A timestamp extraction unit 251, configured to receive the visual rhythm curve and the soundscape rhythm curve, extract the video frame switching timestamp sequence in the visual rhythm curve, and extract the sound pressure level peak timestamp sequence in the soundscape rhythm curve.

[0136] A timestamp normalization unit 252, configured to receive the video frame switching timestamp sequence and the sound pressure level peak timestamp sequence, perform normalization processing on the video frame switching timestamp sequence and the sound pressure level peak timestamp sequence, and generate a normalized timestamp sequence.

[0137] An alignment path calculation unit 253, configured to receive the normalized timestamp sequence, calculate the alignment path distance through the dynamic time warping algorithm according to the normalized timestamp sequence, obtain the minimum alignment path distance, and use the minimum alignment path distance as a quantization index of the matching degree.

[0138] A failure determination unit 254, configured to receive the minimum alignment path distance; if the value of the minimum alignment path distance exceeds a preset distance threshold, determine that the audiovisual collaboration fails and generate an optimization instruction; wherein, the preset distance threshold is a first threshold.

[0139] Optionally, the multisensory data synthesis module 27 includes:

[0140] A landscape element extraction unit 271, configured to receive the video sequence, perform a view frustum analysis on the video sequence, extract the RGB-D data and texture features of the landscape elements within the viewing angle of each node, and generate visual perception data including spatial depth of field information.

[0141] A sound source identification unit 272, configured to receive the sound pressure level attenuation curve, classify and identify the reverberation components of underwater sound, wind blowing through leaves sound, and bird chirping sound according to the spectral characteristics of the sound pressure level attenuation curve, and generate auditory perception data including the sound source type and the azimuth vector.

[0142] A spatio-temporal alignment unit 273 is configured to receive visual perception data and auditory perception data, align the visual perception data and the auditory perception data according to time stamps and space stamps, and generate a standardized perception frame.

[0143] An embodiment of the present application further provides a computer device, including a memory and a processor. The memory stores a computer program, and when the processor executes the computer program, the steps in the foregoing method embodiments are implemented.

[0144] An embodiment of the present application further provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps in the foregoing method embodiments are implemented.

[0145] For the device embodiments, since they basically correspond to the method embodiments, the relevant parts can be referred to the partial descriptions of the method embodiments. The device embodiments described above are only illustrative. The components described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the present disclosure solution. Those of ordinary skill in the art can understand and implement it without creative efforts.

[0146] The above embodiments only represent several implementation manners of the embodiments of the present application. The description is relatively specific and detailed, but it should not be construed as a limitation on the patent scope of the application embodiments. It should be noted that for those of ordinary skill in the art, without departing from the concept of the embodiments of the present application, several modifications and improvements can still be made, and these all belong to the protection scope of the embodiments of the present application.

Claims

1. A garden design method based on garden simulation, characterized in that, The method includes: S1: In the three-dimensional model of the garden, generate multiple tour paths based on the dynamic behavior data of tourists; wherein, each of the tour paths includes multiple consecutive nodes, and each of the nodes includes spatial coordinates, a line-of-sight direction, and a preset stay duration; S2: According to the line-of-sight directions of the nodes in the tour path, generate a continuous video sequence through the dynamic field-of-view chain algorithm; S3: Based on the three-dimensional spatial form of the tour path, construct an environmental acoustic model through the ray tracing method; according to the environmental acoustic model, simulate the sound at the nodes to generate a sound pressure level attenuation curve; wherein, the sound includes water sounds, the sound of the wind blowing through the leaves, and bird calls; S4: Generate a visual rhythm curve by extracting the time interval of video frame switching from the video sequence; generate a soundscape rhythm curve by extracting the sound pressure level peak time points of the sound pressure level attenuation curve from the environmental acoustic model; S5: Calculate the matching degree between the visual rhythm curve and the soundscape rhythm curve through the dynamic time warping algorithm. If the matching degree is lower than the first threshold, it is determined that the audiovisual collaboration fails; S6: If the audiovisual collaboration fails, adjust the line-of-sight direction and the stay duration of the nodes in the tour path, and re-execute S2 to S5 until the matching degree is greater than or equal to the first threshold; S7: Generate multi-sensory space sequence data according to the video sequence and the sound pressure level attenuation curve; wherein, the multi-sensory space sequence data includes standardized perception frames arranged in a spatio-temporal sequence, and each of the standardized perception frames includes visual perception data and auditory perception data corresponding to the nodes; S8: By mapping the multi-sensory space sequence data to a virtual reality interaction platform, design the garden for the dynamic experience of tourists.

2. The method according to claim 1, wherein The S2 includes: S21: According to the line-of-sight directions of the nodes in the tour path, obtain the line-of-sight direction data of the current node; S22: Based on the line-of-sight direction data, obtain the field-of-view range of the current node through the dynamic field-of-view chain algorithm, and calculate the visible area boundary of the field-of-view range; S23: Generate a continuous video sequence frame by frame according to the visible area boundary and the preset stay duration of the node; S24: Calculate the overlap rate of the field-of-view ranges of adjacent nodes. If the overlap rate is lower than the second threshold, insert a transition viewpoint in the field-of-view range and regenerate the video sequence.

3. The method according to claim 1, characterized in that, The S3 includes: S31: Based on the three-dimensional spatial form of the tour path, construct the geometric topology structure of the environmental acoustic model, and extract the reflection surface parameters of the geometric topology structure; wherein, the reflection surface parameters include surface curvature and material type; S32: According to the preset material acoustic parameter library, map the reflection surface parameters to the sound absorption coefficient and scattering rate; S33: Through the ray tracing method, simulate the propagation path of sound waves in the environmental acoustic model based on the sound absorption coefficient and scattering rate, and count the number of reflections and attenuation duration of the propagation path; S34: Generate the sound pressure level attenuation curve according to the number of reflections and the attenuation duration.

4. The method according to claim 1, wherein The S5 includes: S51: Extract the video frame switching timestamp sequence from the visual rhythm curve, and extract the sound pressure level peak timestamp sequence from the soundscape rhythm curve; S52: Normalize the video frame switching timestamp sequence and the sound pressure level peak timestamp sequence to generate a normalized timestamp sequence; S53: According to the normalized timestamp sequence, calculate the alignment path distance through the dynamic time warping algorithm to obtain the minimum alignment path distance, and use the minimum alignment path distance as the quantization index of the matching degree; S54: If the value of the minimum alignment path distance exceeds the preset distance threshold, it is determined that the audiovisual collaboration fails; wherein, the preset distance threshold is the first threshold.

5. The method according to any one of claims 1 to 4, characterized in that The S7 includes: S71: Perform a visual field cone analysis on the video sequence, extract the RGB-D data and texture features of the landscape elements within the viewing angles of each node, and generate the visual perception data including spatial depth information; S72: According to the spectral characteristics of the sound pressure level attenuation curve, classify and identify the reverberation components of the underwater sound, the sound of the wind blowing through the leaves, and the bird calls, and generate the auditory perception data including the sound source type and the azimuth vector; S73: Align the visual perception data and the auditory perception data according to the spatio-temporal stamps to generate the standardized perception frame.

6. A garden design system based on garden simulation, characterized in that, The system includes: A tourist path generation module, configured to generate multiple tour paths in the three-dimensional model of the garden based on the tourist dynamic behavior data; wherein each tour path includes multiple consecutive nodes, and each node includes a spatial coordinate, a line-of-sight direction, and a preset stay duration; A video sequence generation module, configured to receive the tour path, and generate a continuous video sequence through the dynamic visual field chain algorithm according to the line-of-sight direction of the nodes in the tour path; An acoustic modeling module, configured to receive the tour path; construct an environmental acoustic model through the ray tracing method based on the three-dimensional spatial form of the tour path; simulate the sound at the nodes according to the environmental acoustic model to generate a sound pressure level attenuation curve; wherein the sound includes underwater sound, the sound of the wind blowing through the leaves, and bird calls; A rhythm curve generation module, configured to receive the video sequence and the sound pressure level attenuation curve; generate a visual rhythm curve by extracting the time interval of video frame switching from the video sequence; generate a soundscape rhythm curve by extracting the sound pressure level peak time points of the sound pressure level attenuation curve from the environmental acoustic model; An audiovisual collaboration analysis module, configured to receive the visual rhythm curve and the soundscape rhythm curve; calculate the matching degree of the visual rhythm curve and the soundscape rhythm curve through the dynamic time warping algorithm, and if the matching degree is lower than the first threshold, determine that the audiovisual collaboration fails and generate an optimization instruction; A path optimization module, configured to respond to the optimization instruction, adjust the line-of-sight direction and the stay duration of the nodes in the tour path, and sequentially mobilize the video sequence generation module, the acoustic modeling module, and the rhythm curve generation module to re-execute the operations until the matching degree is greater than or equal to the first threshold; A multi-sensory data synthesis module, configured to receive the video sequence and the sound pressure level attenuation curve, and generate multi-sensory spatial sequence data according to the video sequence and the sound pressure level attenuation curve; wherein, the multi-sensory spatial sequence data includes standardized perception frames arranged in a spatio-temporal sequence, and each standardized perception frame contains visual perception data and auditory perception data corresponding to the node. A virtual reality mapping module, configured to receive the multi-sensory spatial sequence data, and design the garden for the dynamic experience of tourists by mapping the multi-sensory spatial sequence data to a virtual reality interaction platform.

7. The system according to claim 6, wherein The video sequence generation module includes: A line-of-sight direction acquisition unit, configured to receive the tour path, and acquire the line-of-sight direction data of the current node according to the line-of-sight direction of the nodes in the tour path. A visible area calculation unit, configured to receive the line-of-sight direction data, and obtain the field of view range of the current node through a dynamic field-of-view chain algorithm based on the line-of-sight direction data, and calculate the visible area boundary of the field of view range. A frame-by-frame rendering unit, configured to receive the visible area boundary and the preset stay duration of the node, and generate a continuous video sequence frame by frame according to the visible area boundary and the preset stay duration of the node. A transition viewpoint insertion unit, configured to receive the visible area boundary, calculate the overlap rate of the field of view ranges of adjacent nodes, and if the overlap rate is lower than a second threshold, insert a transition viewpoint into the field of view range and mobilize the frame-by-frame rendering unit to regenerate the video sequence.

8. The system according to claim 6, characterized in that, The acoustic modeling module includes: A geometric topology construction unit, configured to receive the tour path, construct the geometric topology structure of the environmental acoustic model based on the three-dimensional spatial form of the tour path, and extract the reflection surface parameters of the geometric topology structure; wherein, the reflection surface parameters include surface curvature and material type. An acoustic parameter mapping unit, configured to receive the reflection surface parameters, and map the reflection surface parameters to the sound absorption coefficient and scattering rate according to a preset material acoustic parameter library. A sound wave propagation simulation unit, configured to receive the sound absorption coefficient and scattering rate; simulate the propagation path of sound waves in the environmental acoustic model based on the sound absorption coefficient and scattering rate through the ray tracing method, and count the number of reflections and attenuation duration of the propagation path. A sound pressure level analysis unit, configured to receive the number of reflections and the attenuation duration, and generate the sound pressure level attenuation curve according to the number of reflections and the attenuation duration.

9. The system according to claim 6, characterized in that The audio-visual collaborative analysis module includes: A timestamp extraction unit, configured to receive the visual rhythm curve and the soundscape rhythm curve, extract the video frame switching timestamp sequence in the visual rhythm curve, and extract the sound pressure level peak timestamp sequence in the soundscape rhythm curve. A timestamp normalization unit, configured to receive the video frame switching timestamp sequence and the sound pressure level peak timestamp sequence, and perform normalization processing on the video frame switching timestamp sequence and the sound pressure level peak timestamp sequence to generate a normalized timestamp sequence. An alignment path calculation unit, configured to receive the normalized timestamp sequence, calculate the alignment path distance through a dynamic time warping algorithm according to the normalized timestamp sequence to obtain the minimum alignment path distance, and use the minimum alignment path distance as a quantization index of the matching degree; A failure determination unit, configured to receive the minimum alignment path distance; if the value of the minimum alignment path distance exceeds a preset distance threshold, determine that the audiovisual collaboration fails and generate the optimization instruction; wherein, the preset distance threshold is the first threshold.

10. The system according to any one of claims 6 to 9, characterized in that The multi-sensory data synthesis module includes: A landscape element extraction unit, configured to receive the video sequence, perform a field of view cone analysis on the video sequence, extract the RGB-D data and texture features of the landscape elements within the perspective of each node, and generate the visual perception data including spatial depth information; A sound source recognition unit, configured to receive the sound pressure level attenuation curve, classify and identify the reverberation components of the underwater sound, the sound of the wind blowing through the leaves, and the bird song according to the spectral characteristics of the sound pressure level attenuation curve, and generate the auditory perception data including the sound source type and the azimuth vector; A spatio-temporal alignment unit, configured to receive the visual perception data and the auditory perception data, align the visual perception data and the auditory perception data according to the spatio-temporal stamps, and generate the standardized perception frame.

Images