https://doi.org/10.31449/inf.v48i6.5514                                                      Informatica 48 (2024) 131-140   131 
Analysis of Immersive Virtual Reality Tourism Resources Based on 
Navier-Stokes Equations 
 
Liyan Shen
 
Hangzhou Vocational and Technical College, Hangzhou, Zhejiang, 310018, China 
E-mail: 1999010007@hzvtc.edu.cn 
 
Keywords: Navier-Stokes equations; smoke simulation; immersive virtual reality; tourism simulation system; natural 
feature recognition; 3D registration 
 
Received: December 3, 2023 
The development of virtual resources and scenic spots, attractions, and transportation as well as 
information, rather than just staying in the physical space. There are many virtual reality tourism 
resources in real life. These are natural resources with special characteristics. Natural geographies, 
such as geological landscapes, and visual elements such as colorful and varied textures and colors can 
be attractive and are an essential element in the construction of virtual landscapes, attractions, and 
transport routes. Virtual reality is an emerging field of computer-based, multimedia-based simulation 
and modeling of physical phenomena in natural and social life, which has a wider range of applications 
in immersive virtual reality tourism. This paper introduces a 3D physical virtual reality tourism resource 
analysis system designed based on the Navier-Statics equation software development environment and 
hardware support.  
Povzetek: Analiza turističnih virov z uporabo navidezne resničnosti in enačb Navier-Stokes omogoča 
realistično simulacijo naravnih znamenitosti in poti. 
 
 
 
1  Introduction 
Online travel is becoming increasingly popular among 
travelers as a new kind of leisure as the tourism industry 
grows and individuals focus more on the trip itself. In the 
process of travel, a very important part is the experience. 
Virtual Reality is a virtualized, dynamic, and interactive 
tool for processing online tourism data and information. It 
simulates complex relationships and behavioral patterns 
between humans and nature in real time through 
computers, multimedia, and other means [1-3]. In this 
paper, we will introduce 3D modeling based on Navier-
Stokes equations and virtual reality software, texture 
mapping method and dynamic interactive virtual scene 
design, and other related technologies to show the effect 
of the landscape; and through the construction of the 
relationship between recreationists and attractions to 
simulate the real environment in the process of recreation 
experience to feel the emotional changes [4-5]. 
Fluid simulation theory refers to the simulation of virtual 
reality, the flow field, pressure, and other physical 
quantities in the real environment for analysis, and 
combined with relevant theories and experimental 
methods to reach conclusions [6]. The principle is a new 
theoretical system, it is based on a computer-based 
platform to establish a new system model based on the 
combination of digital signal processing technology and 
intelligent control; it is practical, that the use of virtual 
reality in the study of human brain nerve cells can greatly 
improve people's cognitive ability and perception speed so 
that it can be more easily applied to real life [7], it can 
effectively improve the tourists in the process of travel It 
can effectively enhance the emotional feelings and 
perceptions of tourists during their travels, bringing a huge 
impact on the tourism industry. Collision detection is an 
integral part of the construction of 3D scenes, and the real-
time, accurate detection of collisions between objects can 
enhance the immersion and realism of virtual scenes. 
Collision detection ensures [8,9].  
It ensures that we can move freely around the scene 
without 'walking into the ground' or 'walking through 
walls'. There are two ways of implementing collision 
detection, using colliders and triggers, which are used in 
different situations. Colliders are components, including 
box colliders, ball colliders, mesh colliders, etc., which are 
added to the Game Object to detect collision information 
[10-12]. Triggers, on the other hand, require the Is Trigger 
property to be ticked in the view panel. Triggers can be 
used if you need to detect collisions but don't want the 
physics of the collision to occur or if you just want to 
detect whether an object is passing through a certain area 
of space. For example, triggers are used to detect the 
proximity of a person to a door to control the opening and 
closing of the door, while colliders are often used when a 
ball bounces off a collision with an object. When a 
character is roaming in a virtual tour system, collision 
detection is used in several ways: firstly [13], between the 
character and the ground, as the character is always subject 
to a downward gravitational force, collision detection 
ensures that the character detects a collision with the 
ground and does not continue to fall. The second point is 
between the character and the building, where the 

132   Informatica 48 (2024) 131–140                                                                                    L. Shen 
building's enclosing box is detected when the character is 
close to the building [14], preventing "passing through the 
wall". The third point is the time between the character and 
the surrounding landscape, such as large trees, rocks, etc., 
and the inability to pass through the interior of these 
objects. 
 
2  Related work  
Table 1 provides an overview of the studies investigating 
virtual reality tourism resources, encompassing a 
comprehensive examination of the relevant research 
findings, methodologies, and insights in the field. 
 
 
 
 
 
 
 
 
Table 1: Summary of literature survey 
 
Reference Objective Result Limitations 
[15] The study demonstrated the 
integration of numerical 
simulation tools with a virtual 
reality platform to replicate an 
evolving fire emergency 
scenario. Using a fluid 
dynamic program and the 
building information model, it 
is calculated in real-time. 
Aditionally, a specialized 
program was utilized to 
replicate in real-time the 
dynamics of the crowd in the 
simulated setting as the 
emergency was being 
evacuated. 
The findings demonstrate the 
potential of the suggested virtual 
reality-based system for 
simulating fire emergencies. 
Decision-makers may use it to 
assist in creating emergency 
plans, and firefighters can use it 
as a training tool to practice 
emergency evacuation 
procedures. 
There are obstacles to this 
integration, though, such as 
possible differences in the 
precision of the simulation and 
the behavior of actual fires. 
Furthermore, the virtual 
depiction may fall short of 
accurately capturing the 
complexities of human behavior 
during evacuations, which 
would make it harder to 
forecast the dynamics of actual 
crowds in emergencies. 
 
 
[16] Building suitable coupling 
methods across nonconforming 
interfaces is the primary 
challenge of building 
nonconforming algorithms. In 
this case, we use a decoupled 
interpolation operator-based 
method that is both 
computationally simple and 
efficient. 
The resultant plan is 
conservative element-wise and 
stable concerning entropy. The 
method's accuracy, stability, and 
entropy conservation are 
verified by numerical 
simulations of the isentropic 
vortex and viscous shock 
propagation. 
The given technique 
successfully conserves entropy 
and keeps the system stable, 
although it may not be able to 
handle complicated fluid 
dynamics problems. The 
accuracy and robustness of its 
real-world applications, beyond 
the simulated isentropic vortex 
and viscous shock propagation, 
may be compromised if it 
struggles with extremely 
turbulent flows or complex 
geometries. 
 
[17] The study provided a brief 
overview of ship and 
watercraft modeling and then 
investigated its potential for 
expansion through the use of 
realistic physics. With the help 
of the Unreal Engine, we can 
see how a Landing Craft Utility 
(LCU) operates and interacts 
with near-shore waves in VR, 
creating a hydrodynamic, data-
driven, immersive watercraft 
simulation experience. 
The outcomes validate the 
viability of using DataTables for 
this purpose. We are also 
looking at alternative 
techniques, such as reading 
binary and HDF5 files straight 
from output files, bypassing 
DataTables. 
Nevertheless, when dealing 
with massive datasets, using 
DataTables could cause speed 
and scalability issues. Although 
there may be some performance 
benefits to reading binary and 
HDF5 files directly, there may 
be certain limitations due to 
compatibility difficulties or the 
greater complexity of data 
handling and analysis. 

Analysis of Immersive Virtual Reality Tourism Resources Based…                         Informatica 48 (2024) 131–140   133 
[18] This study set out to create an 
easy-to-use semi-automated 
approach that would allow 
users to better see CFD 
findings and related data in an 
IVE. The process is user-
friendly and supports popular 
CFD software; even 
inexperienced users were able 
to finish it in approximately an 
hour. 
Immersive virtual and 
augmented reality environments 
allow for stereoscopic viewing, 
which may aid in the data 
extraction process from 
computational fluid dynamics 
(CFD) models by making use of 
enhanced depth cues. 
However, problems develop 
when CFD models are not fully 
integrated with immersive 
VR/AR. Data extraction from 
computational fluid dynamics 
simulations may be hindered by 
issues such as a small field of 
vision, the possibility of motion 
sickness, and the requirement 
for high-quality technology, all 
of which can make smooth 
stereoscopic viewing difficult. 
[19] A virtual reality (VR) tool 
called QuickAware was created 
to raise awareness of the 
dangers of fast clay. This 
research details its concept, 
development, and application. 
The tool's development began 
with a co-design strategy, 
which entailed bringing in 
stakeholder experts to work 
together on creating virtual 
reality scenarios and defining 
the content of the VR 
environment. 
The results show that the tool 
might help raise knowledge 
about the dangers of fast clay 
landslides, which is 
encouraging. 
Although the tool can be useful 
in raising awareness of quick 
clay landslides, it is important 
to note that it has significant 
limitations. Several factors 
might restrict its ability to reach 
broad audiences and increase 
common knowledge, including 
accessibility, cultural variances, 
and technology discrepancies. 
[20] The study provided valuable 
insights into the expanding 
field of augmented reality 
technology and set the stage for 
future research and 
advancements, particularly in 
the fields of education and 
outreach. 
While augmented reality 
sandbox technology should 
improve user engagement and 
interaction, the results reveal 
that variables like the lack of fun 
elements reduce its influence. 
Several restrictions make 
augmented reality sandbox 
technology less than ideal for 
improving user engagement and 
involvement. There has to be 
more fun stuff added to 
improve the user experience as 
a whole since things like not 
having any amusing parts make 
it less effective. 
 
 
 
3. The fluid N-S equation 
3.1 Derivation of the equations 
 
Fluid mechanics is based on modern physical theory to 
analyze complex motions and phenomena and to model 
their application in physics. The method is characterized 
by high accuracy, precision, and ease of understanding. It 
can simulate the deformation and flow characteristics 
curves and velocity distribution patterns of various objects 
interacting with each other in a real environment; it can be 
used to study the vibration characteristics caused by 
different shapes, sizes or position changes at the interface 
between the human eye and the ground. From a 
hydrodynamic point of view, the acceleration of a particle 
is then able to be derived from the full differentiation of 
the particle velocity, so Newton's second criterion derives 
the smoke simulation model needed for this paper, as 
shown in the following equation. 
 
 
a =
Du
dt
                                      (1) 
 
m
Du
Dt
= F                                   (2) 
 
A fluid is made up of many tiny particles, which can be 
compressed and arranged into small particles under certain 
conditions. When an object is moving, its gravity will 
change with the inertial force and make it decrease 
continuously. When the acceleration reaches its maximum 
value, it stops moving; after the acceleration reaches its 
peak, it starts to decelerate or move at a constant speed to 
the minimum value to stay and gradually increase the 
energy loss in the process of changing density and size and 
finally stays in the same place. The process of fluid 
movement can be divided into two phases: the first phase 
is the formation of a solid, the second phase is the liquid 
flow and the third phase is the mechanical flow. When we 
use these two different types of virtual reality technology 

134   Informatica 48 (2024) 131–140                                                                                    L. Shen 
together there are several influencing factors: (1) the 
application of virtual reality technology on physical 
models; (2) the trend of interaction and interaction 
between objects in the simulated environment; (3) the 
energy consumption and velocity loss caused by the fluid 
movement process in the simulated environment. Since 
these objects are all based on physical objects, the sum of 
all the forces is superimposed and then substituted into 
equation (2) to obtain the following expressions: 
 
m
Du
Dt
= mg − V ⋅ ∇P + V ⋅ η∇ ⋅ ∇u           (3) 
ρ
Du
Dt
= ρg − ∇p + η∇ ⋅ ∇u                   (4) 
Du
Dt
= g −
1
ρ
∇p +
η
ρ
∇ ⋅ ∇u                    (5) 
 
For a fluid that cannot be compressed, we can obtain the 
following equation. 
 
d
dt
volume (Ω) = ∬ u ⋅ n = 0
aΩ
              (6) 
∂u
∂t
= f −
1
ρ
∇p − (u ⋅ ∇)u                    (7) 
 
3.2 Discrete form of the Two-Dimensional N-
S system of equations 
The two-dimensional N-S system of equations taking into 
account the effects of inertia is expressed as follows. 
Continuity equation. 
 
∂u
∂x
+
∂v
∂y
= 0                                 (8) 
 
Equations of motion. 
 
ρ
∂v
∂t
+ ρ(u
∂v
∂x
+ v
∂v
∂y
) − μ [2
∂
∂y
(
∂v
∂y
) +
∂
∂x
(
∂v
∂x
+
∂u
∂y
)] +
∂P
∂y
− f
y
= 0                  (9) 
 
The boundary conditions involved in this paper are divided 
into two main categories: velocity boundary conditions 
and pressure boundary conditions. t = 0, the velocity 
boundary conditions and pressure boundary conditions are 
 
u ∣
Γ
i
= û, v ∣
Γ
i
= v ̂                          (10) 
p
n
∣
Γ
i
= p ̂
n
                                 (11) 
 
The weighted form of the N-S equation, based on 
equations (8) and (9), can be obtained: 
 
 ∫ w
1
Ω
{
∂u
∂x
+
∂v
∂y
} = 0                       （12 ） 
  
∫ w
1
Ω
{ρ
∂u
∂t
+ ρ(u
∂u
∂x
+ v
∂u
∂y
) − μ [2
∂
∂x
(
∂u
∂x
) +
∂
∂y
(
∂u
∂y
+
∂v
∂x
)] +
∂P
∂x
− f
x
} = 0            （13） 
 
∫ w
2
Ω
{ρ
∂v
∂t
+ ρ(u
∂v
∂x
+ v
∂v
∂y
) − μ [2
∂
∂y
(
∂v
∂y
) +
∂
∂x
(
∂v
∂x
+
∂u
∂y
)] +
∂P
∂y
− f
y
} = 0           （14） 
The above equation gives the weak form of the integral 
under natural boundary conditions. 
∫ w
1
{
∂u
∂x
+
∂v
∂y
}
Ω
= 0                                          (15 ） 
 
 
∫ [ρw
2
∂u
∂t
− ρw
2
(u
∂u
∂x
+ v
∂u
∂y
) + 2μ
∂w
2
∂x
(
∂u
∂x
) +
Ω
μ
∂w
2
∂y
(
∂u
∂y
+
∂v
∂x
) −
∂w
2
∂x
P − w
2
f
x
] dΩ − ∫ w
2
t
x
dΓ = 0
Γ
              
（16） 
 
∫ [ρw
3
∂v
∂t
⋅ ρw
3
(u
∂v
∂x
+ v
∂v
∂y
) + 2μ
∂w
3
∂y
(
∂v
∂y
) +
Ω
μ
∂w
3
∂x
(
∂u
∂y
+
∂v
∂x
) −
∂w
3
∂y
P − w
3
f
y
] dΩ − ∫ w
3
t
y
dΓ = 0
Γ
                        
（17） 
 
Since the volume and surface forces are neglected, and 
therefore are zero, Eq. 
 
t
x
= (2μ
∂v
∂x
− P) n
x
+ μ (
∂u
∂y
+
∂v
∂x
) n
y
    （18 ） 
t
y
= μ (
∂u
∂y
+
∂v
∂x
) n
x
+ (2μ
∂v
∂x
− P) n
y
   （19 ） 
Since the weight function is arbitrary and its essential 
boundary conditions have been determined, the pressure 
term is treated by the penalty function method using 
 
P = η(
∂u
∂x
+
∂v
∂y
)                           (20) 
 
After such treatment, the continuity equation may not 
satisfy the incompressibility assumption, but when η is 
sufficiently large the quantity on the right-hand side of the 
equation is small, so the compressive deformation of the 
fluid is also small and it is considered that it still satisfies 
the incompressibility condition. In the same region, the N-
S equation and the form of the tensor with the penalty 
function applied to treat the pressure are as follows 
 
u
ij,j
+ f
i
= 0                              (21) 
μ (u
i
(η)
− u
i
)
,ij
+ (p
(η)
− p)
,i
= 0        (22) 
 
3.3 Computational cost  
Virtual reality (VR) tourist resource analysis using Navier-
Stokes equations is computationally intensive. To do real-
time rendering, one needs a lot of RAM, powerful 
algorithms, and a powerful computer to handle the 
complex fluid dynamics simulations run in the Navier-
Stokes framework. To achieve a balance between 
computing efficiency and immersive quality, high-fidelity 
virtual reality experiences that faithfully mimic fluid 
behaviors increase the processing burden. Robust 
hardware and optimization methodologies are thus 
required. To provide smooth and lifelike virtual reality 
tourist experiences, it is critical to address these 
computational obstacles. 
 
 

Analysis of Immersive Virtual Reality Tourism Resources Based…                         Informatica 48 (2024) 131–140   135 
3.4 Algorithmic efficiency  
An efficient algorithmic framework is necessary for the 
investigation of Navier-Stokes equations-based immersive 
VR tourist resources. To recover effectiveness, the 
simulation process may be optimized by using parallel 
computing techniques to share the computational burden. 
To further improve the simulation's overall efficiency, 
adaptive mesh refinement techniques can dynamically 
change the grid resolution, directing computing resources 
to where they are most required. Utilizing these 
algorithmic methodologies enhances the solution for 
investigating virtual tourist situations through the use of 
fluid dynamics simulations, making it more robust and 
scalable. 
 
3.5 Challenges faced during implementation 
Several obstacles stand in the way of fully immersive VR 
tourist resources based on Navier-Stokes equations. A lot 
of processing power is required to integrate complicated 
fluid dynamics models into the virtual world, which might 
cause performance problems. Finding the right balance 
between properly modeling fluid behaviors and ensuring 
real-time interactivity is a challenging task that demands 
precise optimization to ensure a smooth user experience. 
When building an interface to explain complicated physics 
in a way that is both easy to grasp and entertaining, it can 
be difficult to strike a balance between scientific precision 
and user-friendly engagement. For immersive virtual 
reality experiences like these to become widely used in the 
tourist industry, certain obstacles must be overcome. 
 
4. Research on roaming control and 
algorithms for immersive virtual 
reality tourism systems 
4.1 Free roaming design 
Free roaming means that the user can freely control the 
direction, perspective route, etc. when roaming through 
various input devices including the mouse and keyboard. 
Users can use external devices to easily and flexibly 
control and operate the roaming of the virtual scenery. 
Using the keyboard "W", "S", "A", "D", spacebar "Z", 
"SHIFT" and the four directional arrows allow the user to 
move forward, backward, leftward, rightward, upward in 
controller angle, downward in controller angle and other 
operations. The mouse can also be used to turn the virtual 
scenery by dragging and to move the controller forward by 
clicking. The process of using the mouse and keyboard is 
to continuously change the coordinate position and angle 
of the controller and continuously update the rendered 
scene through the keyboard, mouse dragging, and mouse 
clicking events. 
 
 
Figure 1: Architecture of free-roaming design 
 
Autorouting is the process of roaming from the user's 
current position along the shortest route to any target point 
in the scene. During the roaming process, the user can also 
control the forward, backward, left, and right turns of the 
controller with the help of the mouse and keyboard. The 
main difference with free roaming is that the forward and 
backward route is already set by the shortest path and the 
user only needs to control forward, backward, left, right, 
or stop. Automatic roaming is used when the user wishes 
to follow the shortest path from the current location to any 
given location. In this way, the user can easily roam the 
scene without having to constantly try and adjust the route, 
while at the same time becoming familiar with and 
understanding the real situation of the scenery, providing 
the user with relevant information for the actual tour of the 
scenery, and increasing the interaction between the user 
and the attraction, enhancing the user's interest in the 
attraction. The key point of automatic wayfinding roaming 
is to find the coordinates and rotation angle of the next 
forward point and to ensure that the path traveled is the 
shortest after a series of point findings.  
 
4.2 Point-First Roaming and optimization 
When roaming according to point priority, that is to say, 
selecting a point on the map, within the map reachable. 
The point on the map is selected so that it passes through 
as many attractions as possible within the range of the 
points that can be reached, while the shortest possible 
distance is required.  
The point at the current location is set as the initial point 
and the location of the selected point is the final point. A 
complete map is generated based on the nodes of the 
attractions on the map, and the optimal path is calculated 
based on the TSP algorithm. If the final point is the initial 
point, the roaming can be carried out directly according to 
the calculated most available path, if the final point is not 
the initial point, the shortest path is generated according to 
the calculated optimal path, and the final point, when the 
generated path is non-loop, i.e. the shortest path from the 
initial point to the final point.  
 
 

136   Informatica 48 (2024) 131–140                                                                                    L. Shen 
With the help of the dynamic programming method to 
solve the TSP algorithm, the basic idea of the algorithm is 
as follows: as shown in Figure 2, v1, v2, v3... As shown in 
Figure 3, v1, v2, v3...v6 represent node 1 to node 6, n is 
the number of nodes, n=6 in the figure. v1 is the starting 
point, as shown in Figure 3, and the distance from v1 to v1 
is 0, if there is no path between two points in the figure, 
then set to infinity, here there is no such path. 
 
Figure 2: Example diagram of the TSP algorithm 
 
A matrix D is used to store the state of the dynamic 
programming, with each row of D representing a node and 
each column representing a subset. This subset contains 
any subset A of the set V of all nodes except v1. 
 
 
Figure 3: Matrix representation of the graph 
 
The optimization idea is roughly as follows: for automatic 
roaming path selection, a system threshold W is set, 
representing the current attraction between hotspots 
exceeding a value of W is considered a congestion 
situation; the number of current hotspot activities N 
recorded by the system is fed back to the console, and if 
N > W, then W is set to infinity. The path is re-planned 
according to the roaming method chosen by the user, and 
the time cost overhead is currently the lowest for the new 
roaming path. 
 
 
 
 
5   System function realization 
5.1 D Scene Loading Analysis Design 
In the virtual tourism system, the terrain is used to indicate 
the height and undulation. Take the ancient town as an 
example, using the natural terrain formed by the hills 
within the town, most of the surface of the town is covered 
by buildings and vegetation, and the overall terrain is not 
very undulating. This paper uses the Unity terrain editor to 
sculpt the terrain in real-time using brush painting to create 
mountains, rivers, flatlands, highs, and lows. After 
creating a new terrain, you can see in the view panel view 
that the terrain object has a terrain script component 
mounted on it. This component contains seven buttons that 
can be used to draw terrain undulations, surface textures, 
or additional details such as trees, grass, rocks, etc. The 
creation process is as follows. 
 
❖ In the project project, create a terrain by opening 
the Terrain -> Create Terrain option in the menu 
bar. 
❖ Adjust the Terrain Width and Terrain Length 
values to 600 in the Terrain review panel view to 
create a 600*600 terrain. 
❖ (3) Paint different areas of the terrain with the 
appropriate texture according to the terrain 
alignment. Here, the mountain areas are painted 
with a grey texture, the surface is mostly 
vegetation so it is painted with a green texture and 
the paths are painted with a yellowish earth color. 
 
Once the terrain has been created, the next step is to add 
plants to the scene to give it a more realistic feel. Set the 
brush size, tree size, and density in the setting panel and 
plant the trees on the ground by left-clicking on the terrain. 
Similarly, plants and flowers can be planted. You can 
remove plants from an area by holding down the Shift key 
and the left mouse button. This method is more suitable for 
situations where a mass painting of vegetation is required 
when we need a random number of trees and flowers to 
appear, to save time and increase efficiency when planting, 
e.g. large forests, clusters of flowers, etc. Because trees 
planted with the Terrain planting tool cannot be 
individually selected for manipulation, they cannot be 
moved twice or zoomed in and out. 
 
5.2 Map Module Design 
In roaming, the navigation and positioning function helps 
us to see where we are in the scene, also known as. This is 
the mini-map navigation technique. The main technique is 
to calculate the ratio of the character's position to the mini-
map and to refresh in real-time. The main technique is to 
calculate the ratio of the character's position to the mini-
map and to refresh the position of the character's logo on 
the mini-map in real-time. The mini-map is created by 
masking areas outside of the mini-map and moving them 
around. Flow of map module design shown in Figure 4. 
The mini-map is created by masking the area outside the 
mini-map so that all that moves is the background behind 

Analysis of Immersive Virtual Reality Tourism Resources Based…                         Informatica 48 (2024) 131–140   137 
it. It can be roughly divided into three parts: firstly, the 
movable  
 
 
 
Figure 4: Architecture of map module design 
 
The first is a movable dynamic map material, which in 
Unity is mainly achieved through the Render Texture 
dynamic rendering. Secondly, create a material based on 
the Render Texture we got in the first step, and write a new 
material to control the shader code for it. The last step is to 
use NGUI to assign the material to a Panel and draw it in 
the top right corner of the screen to see the mini-map. The 
final step is to implement a real-time display and update of 
the mini-map, which determines the position and direction 
of the arrows in the mini-map based on the character's 
position in the overall scene. 
 
5.3  Background sound embedded 
processing 
Audio playback is very important in scene roaming, 
generally speaking. Unity supports four audio formats, 
Aiff, Wav, Mp3, and Ogg, the first two are mainly used for 
short music and the last two for longer sound effects. The 
main purpose of the background music in the virtual tour 
system is to make the tour more relaxing and enjoyable for 
the visitors in the scene, so the background music is chosen 
to be more ancient and elegant, and the principle of audio 
playback in Unity is through a sound source Audio Source, 
and then an audio listener Audio Listener plays the music. 
The implementation process is as follows: an empty Game 
Object named Audio Source is created in the menu bar to 
host the sound source file and provide the music source. 
After the audio object is bound to the component, to 
control the pause, resume, stop, and volume of the audio, 
a script to control this. Take the background music of the 
start screen as an example, and assign the login_audio.cs 
script to the camera object and drag the Audio Resource 
object from the scene to the Music property of the camera 
login_audio.cs script to assign a value to the Music 
variable. Once this is done, run the scene to achieve the 
appropriate functionality. 
 
6   Comparative analysis 
We evaluate the effectiveness of the proposed method N-
S equation with existing methods like MobileNet [21], 
VGG19[21], ResNet50[21], ResNet50 [21]. Among the 
metrics employ for assessment are recall, accuracy, f1-
score, and precision. Table 2 shows the overall outcome of 
proposed methodology. 
 
Table 2: Overall outcomes of the proposed method 
 
Models 
Accurac
y (%) 
Precisi
on (%) 
Rec
all 
(%) 
F1-
Sco
re 
(%) 
MobileNet [21] 98.9 99 98.9 98.9 
VGG19 [21] 96.1 96.2 96.1 96.1 
ResNet50 [21] 93.4 94.3 93.4 93.5 
N-S equation 
[Proposed] 
99.2 99.45 99.3 
99.0
2 
 
6.1  Accuracy  
In VR tourism, accuracy is the practice of creating an 
online space that is as realistic as possible to provide users 
with an engaging and genuine experience. Figure 5 shows 
the results of accuracy. The suggested N-S equation 
outperformed the previous techniques, achieving the 
greatest accuracy of 99.2%. The accuracy of the 
established approaches, such as MobileNet, VGG19, and 
ResNet50, was 98.9%, 96.1%, and 93.4 %. 
 
 
 
Figure 5: Comparison of accuracy 
 
6.2 Precision  
An immersive experience that uses computer-generated 
surroundings to mimic real-world travel is known as 
virtual reality tourism. Through the integration of state-of-
the-art technology, it facilitates the virtual exploration of 

138   Informatica 48 (2024) 131–140                                                                                    L. Shen 
many places. Precision in sensory reproduction and user 
engagement is a measure of precision in virtual reality 
tourism. Figure 6 shows the results of precision. The 
suggested N-S equation outperformed the previous 
techniques, achieving the most incredible precision of 
99.45%. The accuracy of the established approaches, such 
as MobileNet, VGG19, and ResNet50, was 99.0%, 96.2%, 
and 94.3% 
 
 
Figure 6: Comparison of precision 
 
6.2 Recall 
In the context of VR tourism, recall is a measure of how 
well the virtual world captures meaningful experiences, 
suggesting that all significant characteristics of a place are 
included. The formula for it is the proportion of favorable 
experiences that were positive relative to the total of 
pleased and negative events. Figure 7 shows the results of 
recall. The suggested N-S equation outperformed the 
previous techniques, achieving the best recollection of 
99.3%. The memory of the established approaches, such as 
MobileNet, VGG19, and ResNet50, was 98.9%, 96.1%, 
and 93.4% 
 
Figure 7: Comparison of Recall 
 
6.3 F1-Score 
When evaluating the thoroughness and correctness of a 
virtual tourist experience, the F1-score is the harmonic 
mean of recall and precision. Figure 8 shows the results of 
the F1 score. The suggested N-S equation outperformed 
the previous techniques, achieving the maximum F1 score 
of 99.2%. The F1 score of the established approaches, such 
as MobileNet, VGG19 and ResNet50, was 98.9%, 96.1% 
and 93.5%  
 
Figure 8: Comparison of F1-score 
 
7  Discussion  
Investigating the impact of VR on the travel industry is the 
main aspect of studying immersive VR tourism. As part of 
this process, we assess user experiences, emerging 
technologies, and their potential to transform the travel 
industry by offering a novel and captivating means for 
individuals to discover and engage with other locations. 
MobileNet [21] is that it has to choose between being 
accurate and being computationally efficient. Its 
lightweight architecture makes it less accurate than more 
sophisticated models, which makes it less suitable for 
applications that need high-level accuracy and fine-
grained feature extraction, even if it is optimized for 
mobile devices. The deep design of VGG19 [21] makes it 
extremely computationally and memory intensive, which 
makes it less practical for real-time applications or devices 
with limited resources, even though it does a great job at 
feature representation. ResNet's [21] great performance in 
DNNs isn't without its drawbacks, though. Its deep design 
increases computational cost, and there's a chance of 
disappearing gradients and overfitting on smaller datasets. 
To optimize its performance, careful tweaking is required. 
By finding a happy medium between precision and 
computing efficiency, our suggested N_S equation 
approach overcomes shortcomings of popular models such 
as ResNet, MobileNet, and VGG19. It provides a more 
effective answer for immersive VR tourist applications by 
reducing the impact of problems like overfitting, vanishing 
gradients, and calculations that are heavy on system 
resources. 
 
8  Conclusion 
This paper designs and implements an efficient, highly 
portable, and scalable Unity3D-based virtual tour system. 
The system is based on the simulation function of the N-S 
equation and the Unity3D engine. No additional plug-ins 
are required, and integration between modules is achieved 
through XML configuration files, which can be configured 
by the user. For the bottleneck of 3D scenes in the virtual 
tourism system, the article optimizes the design of material 
files and model files for the construction of 3D scenes and 
the steps of construction to achieve an improved loading 

Analysis of Immersive Virtual Reality Tourism Resources Based…                         Informatica 48 (2024) 131–140   139 
and performance display of 3D scenes. The system 
supports a variety of media methods, including 3D objects, 
pictures, audio and video, etc. The system's colorful 3D 
scenes (including undulating surfaces, rich and lush 
vegetation, and highly distinctive buildings) do not use 
traditional 3D GIS technology but the Unity engine, which 
has more mature technology in terms of graphics, physics, 
interface, and scripting. The system has information query 
and storage functions, adds roaming background music, 
makes the virtual tour more vivid, through the particle 
system more realistic simulation of the water flow effect 
and waterfall effect, and realizes the user mouse-free way 
finding and fixed-point way finding function. The 
proposed method shows significant results when compare 
to traditional approaches in terms of accuracy (99.2 %.), 
precision (99.45 %), recall (99.3 %) and f1-score 
(99.02 %). The system aims to improve China's tourism 
information development by providing a comprehensive 
and realistic display of scenic spot information. It is 
affordable to build and maintain, and its strong portability 
contributes to the optimization of service capacity and 
resource allocation. This system enhances user experience 
and provides practical information for tourists, enhancing 
the overall tourism experience in China.  
 
References 
[1] Voronkova, L.P., 2018, December. Virtual tourism: 
on the way to the digital economy. In IOP Conference 
Series: Materials Science and Engineering (Vol. 463, 
p. 042096). IOP Publishing.  
[2] Iacovino, A., De Paolis, L.T. and Ndou, V., 2020. 
Technologies to Support Tourism Innovation and 
Cultural Heritage: Development of an Immersive 
Virtual Reality Application. In Augmented Reality, 
Virtual Reality, and Computer Graphics: 7th 
International Conference, AVR 2020, Lecce, Italy, 
September 7–10, 2020, Proceedings, Part II 7 (pp. 3-
14). Springer International Publishing.  
[3] Neuburger, L., Beck, J. and Egger, R., 2018. The 
‘Phygital’tourist experience: The use of augmented 
and virtual reality in destination marketing. In 
Tourism planning and destination marketing (pp. 
183-202). Emerald Publishing Limited.  
[4] Choi, N. and Sung, M., 2024. CWD-Sim: Real-Time 
Simulation on Grass Swaying with Controllable 
Wind Dynamics. Applied Sciences, 14(2), p.548.  
[5] Wang, X., Lv, M., Liu, S., Li, J., Zhang, J. and Meng, 
F., 2023. An open GIS based 3D simulation software 
to predict cooling tower drift diffusion. Scientific 
Reports, 13(1), p.18186. 
[6] Boettcher, K.E. and Behr, A.S., 2021. Using virtual 
reality for teaching the derivation of conservation 
laws in fluid mechanics. Int. J. Eng. Pedagog., 11(4), 
pp.42-57. 
[7] Wang, B., Hu, S.J., Sun, L. and Freiheit, T., 2020. 
Intelligent welding system technologies: State-of-
the-art review and perspectives. Journal of 
Manufacturing Systems, 56, pp.373-39.  
[8] Ye Zanwei. Applicationofmeshlessmethods in 
thenumericalsimulationoftransonic and 
supersonicflowfields [D]. Nanjing 
UniversityofScience and Technology, 2017. 
[9] Sun Hui, Tan 
Junjie.Applicationofmeshlessalgorithmfor Euler 
equation in compressibleflowfields[J]. 
JournalofBallisticArrows and 
Guidance,2018(02):171-174.  
[10] Tamminen, V., 2022. Combat hit detection system for 
3D action role-playing games in Unity.  
[11] Strid, J., 2023. Signed Distance Field For Deformable 
Terrain Shovel Collision Detection. 
[12] Jindariani, S., Meloni, F., Pastrone, N., Aimè, C., 
Bartosik, N., Barzi, E., Bertolin, A., Braghieri, A., 
Buonincontri, L., Calzaferri, S. and Casarsa, M., 
2022. Promising technologies and R&D directions 
for the future muon collider detectors. arXiv preprint 
arXiv:2203.07224.  
[13] Sridhar, V., Podjaski, F., Alapan, Y., Kröger, J., 
Grunenberg, L., Kishore, V., Lotsch, B.V. and Sitti, 
M., 2022. Light-driven carbon nitride 
microswimmers with propulsion in biological and 
ionic media and responsive on-demand drug delivery. 
Science robotics, 7(62), p. eabm1421.  
[14] Haghighatgou, N., Daniel, S. and Badard, T., 2022. 
A method for automatic identification of openings in 
buildings facades based on mobile LiDAR point 
clouds for assessing impacts of floodings. 
International Journal of Applied Earth Observation 
and Geoinformation, 108, p.102757. 
[15] Lorusso, P., De Iuliis, M., Marasco, S., Domaneschi, 
M., Cimellaro, G.P. and Villa, V., 2022. Fire 
emergency evacuation from a school building using 
an evolutionary virtual reality platform. Buildings, 
12(2), p.223. 
[16] Fernández, D.C.D.R., Carpenter, M.H., Dalcin, L., 
Zampini, S. and Parsani, M., 2020. Entropy stable 
h/p-nonconforming discretization with the 
summation-by-parts property for the compressible 
Euler and Navier–Stokes equations. SN Partial 
Differential Equations and Applications, 1(2), p.9. 
[17] Larico, J.A.T., 2020. Towards interactive simulation 
of soft, rigid and viscous objects in immersive virtual 
reality. 
[18] Venn, J., Larkee, C.E., Garcia, G.J., Rayz, V.L. and 
LaDisa Jr, J.F., 2023. A workflow for viewing 
biomedical computational fluid dynamics results and 
corresponding data within virtual and augmented 
reality environments. Frontiers in Medical 
Technology, 5, p.1096289. 
[19] Alene, G.H., Depina, I., Thakur, V., Perkis, A. and 
Bruland, O., 2023. QuickAware: a virtual reality tool 
for quick clay landslide hazard awareness. Natural 
Hazards, pp.1-30. 
[20] Cranmer, E.E., tom Dieck, M.C. and Fountoulaki, P., 
2020. Exploring the value of augmented reality for 
tourism. Tourism Management Perspectives, 35, 
p.100672. 
[21] Chhikara, P., Kuhar, H., Goyal, A. and Sharma, C., 
2023, January. DIGITOUR: Automatic Digital Tours 
for Real-Estate Properties. In Proceedings of the 6th 
Joint International Conference on Data Science & 

140   Informatica 48 (2024) 131–140                                                                                    L. Shen 
Management of Data (10th ACM IKDD CODS and 
28th COMAD) (pp. 223-227).