Available via license: CC BY 4.0
Content may be subject to copyright.
Applied Mathematics and Nonlinear Sciences, 9(1) (2024) 1-16
Applied Mathematics and Nonlinear Sciences
https://www.sciendo.com
†Corresponding author.
Email address: Lixiaona1799@126.com
ISSN 2444-8656
https://doi.org/10.2478/amns-2024-0770
© 2023 Xiaona Li, published by Sciendo.
This work is licensed under the Creative Commons Attribution alone 4.0 License.
Cultural Inheritance and Artistic Construction of Non-heritage Dance in Northern
Anhui in the Era of Artificial Intelligence
Xiaona Li1,†
1. Suzhou University, Suzhou, Anhui, 234000, China.
Submission Info
Communicated by Z. Sabir
Received January 31, 2024
Accepted February 7, 2024
Available online April 1, 2024
Abstract
This paper explores the integration of image processing, motion capture, and virtual reality technologies to digitize and
visualize dance. We capture the core dynamics of dance movements by extracting key frames and movement features
from dance videos. Our analysis of motion capture data, exemplified by the “Flower Drum Lantern” dance, reveals a
maximum vertical foot displacement of 72 cm and hip displacement of 93 cm. Virtual display technology significantly
enhances the visual representation and dissemination of dance performances. This innovative approach to documenting
and showcasing dance not only aids in preserving and transmitting intangible cultural heritage but also boosts public
awareness and appreciation for such heritage.
Keywords: Cultural Heritage; Artistic Construction; Artificial Intelligence; Virtual Presentation.
AMS 2010 codes: 97N80
Xiaona Li. Applied Mathematics and Nonlinear Sciences, 9(1) (2024) 1-16
2
1 Introduction
Intangible cultural heritage refers to all kinds of traditional cultural expressions that have been handed
down from generation to generation by people of all ethnic groups and regarded as part of their
cultural heritage, as well as physical objects and places related to traditional cultural expressions,
which mainly include ten categories: folk literature, traditional fine arts, traditional arts and crafts,
traditional music, traditional dances, traditional dramas, traditional medicines, music and arts,
folklore and traditional sports, and amusement arts and acrobatics. As a unique form of artistic
expression, non-heritage dance combines national and regional characteristics, carries rich historical
memories and cultural connotations, and is a treasure of Chinese traditional culture [1-3]. As a kind
of cultural wealth, non-legacy dance is an art form created by people of different regions and
nationalities in their long-term life and production practice, and it is an important part of the
traditional culture of the Chinese nation. These dances have a long history and have been precipitated
for thousands of years, carrying rich historical and cultural connotations. China is a vast country, and
different regional and ethnic characteristics have created diverse and unique non-heritage dance forms
[4-6].
In the new period, under the environment of rapid development of science and technology and
increasingly frequent globalization and communication, the inheritance and development of non-
heritage dance are also facing new opportunities and challenges. How to protect the traditional
essence of non-heritage dance in the context of modern society and, at the same time, inject new
elements into it so that it can be revitalized in contemporary life has become an important content of
current research [7-9].
Folk dance is a splendid treasure of Chinese civilization, carrying rich historical connotations and
national spirit and possessing great cultural charm. However, in the process of modernization, the
inheritance and development of non-heritage dances are facing many challenges, and it is imperative
to find a development path that meets the requirements of the new period [10]. Based on the analysis
of black dance archives, Carr, J. et al. revealed that public attention to the dispersed black dance
cultural heritage is a form of intangible dance cultural heritage preservation and transmission to
prevent it from being forgotten by future generations [11]. They pointed out that the research on AI
in China has replaced the old “mechanical reductionist” methodology characterized by “divide and
conquer” with the information ecology approach of “integration and growth” [12]. Hu, M. et al.
developed a Bayesian network-based teacher model to evaluate different stages of dance training,
pointing out that inclusive dance helps to improve human coordination and also improves physical
condition [13]. Peng, H et al. proposed a mechanism for automatic aesthetic evaluation of robotic
dance movements based on the integration of multiple visual features. After simulation experiments,
it was corroborated that the evaluation accuracy of the mechanism reached 75%, which was better
than the traditional method [14]. Ozcimder, K. et al. explored the traditional performing culture and
art forms of salsa and evaluated them with an AI jury and a professional jury, and the comparative
results confirmed the reliability of the AI jury [15]. Zhuang, X. et al. compared the impacts of tourism
development on the socio-cultural aspects of three ancient villages based on participatory in-depth
interviews while examining the factors that contributed to the change in the moral values of the
residents and found that tourism development is the key to the residents’ moral value change [16].
Candela, G. et al. proposed a hybrid algorithm to optimize and refine Web services (WS), aiming to
improve the impact and discoverability of cultural heritage [17]. Karadeniz, C. et al. elaborated that
culture is a complex of knowledge, beliefs, arts, morals, and customs that come together and that the
purpose of museum The purpose of education is to understand the culture, internalize its similarities
and differences and provide awareness of cultural heritage preservation, and talked about Ankara
University, museum displays, and teaching on the theme of cultural heritage [18].
Cultural Inheritance and Artistic Construction of Non-heritage Dance in Northern Anhui in the Era of
Artificial Intelligence
3
The article takes the non-heritage dance of northern Anhui as the research object, and the image
processing technology extracts the key features of the dance video and applies the motion capture
technology to record and analyze the dance movements in detail. The use of virtual reality technology
for three-dimensional reproduction and display enhances the expressiveness and viewability of the
dance. The communication effects of dance are analyzed through social media platforms to explore
the role of modern technology in traditional cultural inheritance.
2 Feature Extraction of Non-legacy Dance Video Resources
2.1 Dance video image manipulation
2.1.1 Bitmap Structure Analysis
A bitmap is divided into device-independent bitmap and device-related bitmap. Device-independent
bitmap (DIB) can be copied and disseminated and can be used for different systems, while the device-
related bitmap (DDB) is created and processed in memory, it does not come with a color palette; its
color is determined by the device that displays it, it is device-related.
DIB file consists of four parts: bitmap file header, bitmap information header, color table entries, and
bitmap image data, of which the first three are managed by the data structures
BITMAPFILEHEADER, BITMAPINFOHEADER and RGBQUAD, respectively, while
BITMAPINFOHEADER and RGBQUAD structure using the BITMAPINFO combination.
2.1.2 Bitmap process
Bitmap processing techniques and algorithms include bitmap processing, regional processing,
morphological processing, inter-image operations (synthesis, algebraic operations, logical operations,
image fusion, etc.)., geometric transformations, and so on. Geometric transformations and other
related things.
Establish the contrast processing class Contrast Process to adjust the contrast of the image. To change
the contrast of the image, it is necessary to change the gray levels of red, green, and blue in a consistent
manner. Assuming that the contrast of the image to be expanded by
( )
0nn
units, then the uniform
transformation formula for the three components of red, green and blue is.
0,
255, 255
255
( ) ,
255 2
fn
g f n
f n otherwise
n
= −
−−
(1)
If the contrast is to be reduced by
( )
0nn
units, the transformation formula for the three
components is.
255 2
255
n
g f n
+
=−
(2)
Xiaona Li. Applied Mathematics and Nonlinear Sciences, 9(1) (2024) 1-16
4
f is the original value; g is the changed value. The three component values here are in the range of
0~255, beyond which they will be truncated.
Establish the grayscale processing class CGrayProcess, which will deal with grayscale linear
transformation, histogram equalization, histogram specification, and obtaining the average value of
grayscale, etc. The color image is converted to grayscale using the following formula.
0.3 0.59 0.11gray r g b= + +
(3)
,,r g b
represent red, green, and blue, three color values,
gray
is the gray value.
The equalization algorithm for histograms is presented here. The probability that 256 gray levels do
not occur is.
, 0,1, ,255
k
k
n
pk
n
= =
(4)
n
is the total number of pixels,
k
p
is the frequency of occurrence of the
k
th gray level. Then let
the cumulative distribution function
q
be expressed as.
,
0
, 1,2, ,255
k
kr
q p k
=
= =
(5)
In this way, we can establish the relationship between gray levels and mapping.
255
k
kq→
(6)
The original meter gray level of
k
pixels into
255
k
q
, this conversion is the histogram
equalization algorithm.
Establishment of the image shading class CRender Process, which is used to generate grayscale
images with specified color shading. This has important applications in certain applications, such as
atmospheric cloud processing and medical image processing.
To establish a photographic-based exposure image processing class CSolarization Process, the
exposure image algorithm is to invert the r, g, and b components whose values are less than a certain
constant, and the magnitude of this constant can affect the degree of exposure.
2.1.3 Bitmap Geometry Transformations
Bitmap geometric transformation can change the spatial location of objects in the image, using it to
realize image geometry correction, image alignment, style transformation, deformation, and other
operations.
Image geometric transformations include translation, scaling, rotation, mirroring, and so on. In this
process, we have to deal with the interpolation algorithm as if the pixel, which is the core algorithm
of image geometric operations, because when the spatial transformation of the role of pixels in the
integer coordinates, non-integer coordinates will be generated. Generally, the methods used to deal
with them include nearest-neighbor interpolation, bilinear interpolation, and triple convolution
Cultural Inheritance and Artistic Construction of Non-heritage Dance in Northern Anhui in the Era of
Artificial Intelligence
5
methods, among others. The method of nearest-neighbor sampling is a simple non-integer rounding,
and it is clear that the obtained image quality is not high. The bilinear interpolation method is not
good. The method of bilinear interpolation is not very good. The image quality is not high. Bilinear
interpolation is like this. Assuming that the transformed floating point coordinates are
( )
,i u j v++
,
where
i
and
j
are non-negative integers, and
u
and
v
are floating point numbers in the interval
[0,1]
, the destination pixel value is obtained by co-computing the four pixels in the original image,
i.e.
( , ) (1 )(1 ) ( , ) (1 ) ( , 1)
(1 ) ( , ) ( 1, 1)
f i u j v u v f i j u vf i j
u v f i l j uvf i j
+ + = − − + − +
+ − + + + +
(7)
The image quality obtained by the bilinear bracketing method is higher, but it has the nature of low-
pass, which will make the image wheel modulus and the third convolution method is the most accurate
method, but the computational volume is also the largest, which considers the 16 neighboring points
around the
( )
,i u j v++
, and the target value of the pixel is obtained by the following formula.
( , ) [ ][ ][ ]f i u j v A B C+ + =
(8)
Of which.
( ) ( ) ( ) ( )
( )
1 1 2A S u S u S u S u= + − −
(9)
( ) ( ) ( ) ( )
( )
1 1 2 T
C S v S v S v S v= + − −
(10)
( 1, 1) ( 1, ) ( 1, 1) ( 1, 2)
( , 1) ( , ) ( , 1) ( , 2)
( 1, 1) ( 1, ) ( 1, 1) ( 1, 2)
( 2, 1) ( 2, ) ( 2, 1) ( 2, 2)
f i j f i j f i j f i j
f i j f i j f i j f i j
Bf i j f i j f i j f i j
f i j f i j f i j f i j
− − − − + − +
− + +
=
+ − + + + + +
+ − + + + + +
(11)
23
23
1 2 | | | | ,| 1|
( ) 4 8| | 5 | | | | ,1 | | 2,
0,| 2 |
sin( )
()'
S
hIt s for aS Appro c
− +
= − + −
=
(12)
The geometric transformation of bitmaps in the video library system also plays an important role.
Such as the different frame sizes of the video, we can first decode into a sequence of images after
geometric transformation and then compress so that the two video frames can be the same size of the
frame, can be connected, and have a variety of operations. Other roles do not individually state the
role of the ClmageGeometry class in achieving geometric transformation operations of images.
2.2 Dance Video Keyframe Extraction
Music and dance video movements are complex and varied, and the problem of repetitive movements
can cause difficulties in the analysis and recognition of dance movements. Therefore, extracting the
Xiaona Li. Applied Mathematics and Nonlinear Sciences, 9(1) (2024) 1-16
6
keyframes of music and dance videos will be an effective solution to improve the rate of dance video
analysis and reduce computational complexity. The keyframes are the image frames that have the
least redundancy in the video and are a representation of the video. In the case of the dance video, the
extraction of the keyframes is the process of removing the redundancy, and at a certain point in time
when the music has a strong performance, and the current frame has a large difference from the
previous keyframe, the image frame will be recognized as a new keyframe.
2.2.1 Key Frame Extraction Algorithm Ideas
In order to extract a set of keyframes with less redundancy and to summarize the content of the video,
this paper will carry out optical flow computation on the image sequences of the dance action video
after frame splitting, which is able to match the movements with large displacements and estimate
the optical flow of the smaller objects as well. The main idea of the optical flow computation is shown
in Figure 1. The main idea of the optical flow calculation method is shown in Fig.1.The optical flow
map reflects the movement direction and speed of the dance action, and the entropy calculation is to
count the information of each optical flow image. However, for music and dance videos, the music
features are also very important. Therefore, in this paper, we will extract the envelope features and
energy features of the music and merge these features with the entropy sequence to obtain a music-
related entropy sequence. Finally, this paper will extract music features larger than the entropy
sequence and fuse them with the entropy sequence. The key frame is the frame that is larger than the
threshold value, and the collection of keyframes is the video summary of the dance video.
Video Image
Sequences Optical Flow
Matching
Calculation
Entropy
Calculation
Music Feature
Extraction
Feature
Fusion Key frame
extraction
Audio
Figure 1. The main idea of the method of light flow
2.2.2 Dance Motion Feature Extraction
The traditional optical flow algorithm will be limited in some moments, although it can track the
target in the image. However, for the discontinuous movement, occlusion phenomenon, and large
displacement problems, the optical flow calculation effect is not very good. Therefore, through
continuous analysis of the problem, process, and exploration of the corresponding solution, the optical
flow method will be targeted. For the dance video, it is possible to accurately calculate each change
in the dance movement and not miss the changes in the movements of those small limbs. The method
for calculating optical flow is shown in the formula.
( ) ( )
11
( ) ( ) ( ) ( )
,
color grad smooth
match desc
E w E w E w E w
E w w E w
= + +
++
(13)
( )
2
21
( ) I ( ( )) I ( )
color
E w x w x x dx
= + −
(14)
Cultural Inheritance and Artistic Construction of Non-heritage Dance in Northern Anhui in the Era of
Artificial Intelligence
7
( )
2
21
( ) I ( ( )) I ( )
grad
E w x w x x dx
= + −
(15)
( )
22
( ) | ( )| | v( ) |
smooth
E w x x dx
= +
(16)
( )
2
1
( ) ( ) ( ) ( ) ( )
match
E w x x w x w x dx
= −
(17)
( ) ( )
2
1 2 1 1
( ) ( ) ( )
desc
E w x f x w x f x dx
= + −
(18)
Among them,
,
and
are the adjustable weight parameters.
( )
color
Ew
is the assumption of
luminance invariance, which is applicable to both color and grayscale images. The effect of
illumination is unavoidable, so in order to reduce the effect of illumination, it is necessary to add the
gradient constraint
( )
grad
Ew
on top of it. And then
( )
smooth
Ew
soothes it. The last two items are to
construct the descriptor matching and find the minimum value of it using the multivariate model and
optimization. The effect of calculating the optical flow of the two neighboring frames is illustrated.
3 Virtual presentation of non-heritage dances
The dance performance data obtained by motion capture accurately records the length of the main
joints of the actor’s body, as well as the angle of joint movement. However, the viewer is unable to
obtain an intuitive understanding of the dance performance through these data alone, and for this
reason, it is also necessary to reproduce the original performance through virtual display technology.
3.1 Representation of Human Movement
In motion capture and character animation, the skeletal structure of the human body is usually
represented as a chain of bones, which are rigidly connected to each other, and the child’s bones are
affected by the movement of the parent bones. The motion file should record not only the length of
the bones but also the position and orientation of each bone in each frame. The most common BVH
format shows that the record of each bone’s position and orientation has 6 degrees of freedom. The
Euler angles of the bone’s displacement and rotation along the
X
,
Y
,
Z
coordinate axes,
respectively. let
x
R
,
y
R
,
z
R
be the rotation matrices of the three
33
‘s corresponding to the Euler
angles, and
x
T
,
y
T
,
z
T
be the displacements in the direction of the three coordinate axes, then the
rotation matrices of the bone are.
11 12 13
21 22 23
31 32 33
0 0 0 1
x
y
z
R R R T
R R R T
MR R R T
=
(19)
Of which.
33 x y z
R R R R
=
(20)
Xiaona Li. Applied Mathematics and Nonlinear Sciences, 9(1) (2024) 1-16
8
Formula (19) is the transformation matrix in the local coordinate system of a bone, and for bones
other than the root bone, it is necessary to calculate the sum of the local transformation matrices of
all the parent bones, e.g., if the vector representing the length and the initial orientation of the bone
of the right foot in the local coordinate system of the bone of the right foot is A, the spatial position
of the right foot during the motion is.
Hips RighhHip RighKnee Righhnukle
A M M M M A=
(21)
3.2 Skeleton skinning animation implementation
3.2.1 Calculation of vertex positions in the skinned mesh
As described in section 3.1, the transformation matrix
i
M
of each bone in the current frame can be
obtained from the degrees of freedom of each bone in each frame recorded in the action file, and the
transformation matrix of each bone in the current frame can be obtained from the transformation
matrix of each bone in the localized coordinate system, and to obtain the transformation matrix of a
bone in the world coordinate system, it is also necessary to consider the parent bone of the bone, and
the transformation matrix of each bone in the world coordinate system is obtained by setting up the
skeleton
i
as a child skeleton of the skeleton
1i−
, as follows.
1i i i
T T M
−
=
(22)
Where
0
T
is the unit matrix.
Since the coordinates of the vertices in the skin mesh are computed according to the world coordinate
system, the spatial position of the skin vertices in each frame is computed by first transforming the
vertices to the coordinate system of the skeleton, and this transformation matrix
i
O
is easily obtained
from the initial position information of the skeletal chain. The initial position information of the bone
chain can easily yield the transformation matrix, and we can obtain the final transformation matrix as
follows.
t i i
F TO=
(23)
The position of the vertex P is driven by the skeleton at the time of motion.
i
P F P=
(24)
The above equation only considers the case where the vertex is driven by a single bone, and the shape
of the skin is generally affected by multiple bones at the joints of the human body, and the animation
realized by Eq. (24) has wrinkles at the joints even though the moving skin is seamless, which is due
to the fact that the position of the vertex is driven by a single bone, thus leading to the wrinkled effect.
To solve the above problem, a common solution is to use the vertex blending technique, i.e.assigning
multiple bones and weights to a vertex. Let a vertex be affected by a total of k bones, which can be
easily obtained from Eq. (24).
1
k
jj
j
P w F P
=
=
(25)
Cultural Inheritance and Artistic Construction of Non-heritage Dance in Northern Anhui in the Era of
Artificial Intelligence
9
Of which.
1
1
k
j
j
w
=
=
(26)
j
w
in Eq. (26) is generally calculated out of the modeling software or set manually by the animator.
As you can see, the folds in the joints have been significantly eliminated.
3.2.2 Accelerated drawing based on Vertex Shader
In the traditional fixed-function graphics pipeline, we need to calculate the position
P
, of all vertices
after transformation by CPU, which is a large computational overhead. Usually, a performance scene
contains multiple character models, and a character model has thousands to tens of thousands of
vertices, which makes real-time display of animation difficult. This makes it difficult to display
animation in real time.
In order to improve the rendering efficiency, we use the programmable rendering pipeline of the
graphics card to transfer the Calculation of the new position of the vertex to the Vertex Shader in each
frame and utilize the powerful matrix motion function of the GPU to improve the rendering efficiency.
The Vertex Shader’s rendering flow is shown in Figure 2.
Raw vertex
data Viewpoint
transformation Projection
transformations Rasterisation
World Transform
Dynamically
calculates new
vertex positions
during animation Vertex Shader
GPU
Figure 2. The rendering process for Vertex Shader
4 Characteristics of non-heritage dances in northern Anhui
Northern Anhui non-heritage dance is an important cultural style in the northern Anhui region. The
northern Anhui non-heritage dance has a unique artistic charm and deep cultural connotations. It is
our mission to protect and inherit the northern Anhui non-heritage dance, which has been given to us
by the times. Northern Anhui is a geographic and cultural concept. The Yangtze River and the Huaihe
River run through Anhui, which divides Anhui into the southern part of Anhui, the Jianghuai River,
and the north of Anhui. Northern Anhui includes six municipalities, including Bengbu, Fuyang,
Bozhou, and others. Northern Anhui is the most popular dance in the northern Anhui region,
combining the colors of the region and the charm of ethnicity, folklore, and folk culture. The Northern
Anhui non-heritage dance is widely popular in northern Anhui. It has the color of the northern Anhui
region, as well as the charm of nationalities, folklore, and folk culture.
The characteristics of non-heritage dances in northern Anhui are mainly reflected in multi-oceanic,
regional, and inclusiveness. In this paper, we take the “Flower Drum Lantern” dance as a research
Xiaona Li. Applied Mathematics and Nonlinear Sciences, 9(1) (2024) 1-16
10
case and use image manipulation and key frame extraction to obtain the music and rhythmic
characteristics of the dance video resources.
4.1 Characteristics of dance movements of non-heritage dances in Northern Anhui Province
After collecting all the characteristic dance units in the original movement library of the Northern
Anhui non-heritage dance movements, we determined the signature movement set of the style based
on the frequency of the body joint movements in these movement units. The overall frequency of the
key movements of the non-heritage dances is shown in Table 1. The dance combination in the dances
consists of more than two dance movements that are connected together, and the movements are
composed of local movement characteristics of four major joints, namely the head, upper limbs, lower
limbs, and trunk, and then combined into a whole body dance movement. We randomly picked 80
dance movements that are characteristic of the Northern Anhui style’s movement categories. We
randomly selected 80 characteristic dance units in the movement category of the northern Anhui style
and combined them into a signature movement of northern Anhui NRD dance according to the
frequency of the movement of the joints of these body parts, among which “crossing the waist with
both hands” and “single movement” are the key movements of northern Anhui NRD dance. Among
them, the number of times that “folding the waist with both hands” and “kneeling with one leg”
appeared was zero, reflecting the movement inertia of the Northern Anhui non-heritage dance.
Table 1. The overall frequency of the key movements of the dance
Body part
Action
Occurrence number
Frequency
Head
Head high
801
10.01
Bow
54
0.68
Flat sight
256
3.2
Upper limb
Torso
Torsional waist
332
4.15
Trying waist
444
5.55
Straighten up
212
2.65
Tappet
618
7.73
Chest
29
0.36
Chest turn
234
2.93
Arm
Single arm
767
9.59
Hand lift
504
6.3
Hands on hips
0
0
One hand on hip
71
0.89
Cross over
80
1
Natural droop
53
0.66
Lower limb
Muddle
196
2.45
Bend knees
403
5.04
peacocks
444
5.55
kneel on two legs
299
3.74
Kneel on one leg
0
0
For the dance movement of “Flower Drum Lantern” selected in this paper, its rhythm is mainly
reflected by the jumping of the feet and the squatting of the arms. Figure 3 shows the changes in the
Cultural Inheritance and Artistic Construction of Non-heritage Dance in Northern Anhui in the Era of
Artificial Intelligence
11
dance postures of “Flower Drum Lantern,” (a) is the trajectory of the left foot’s vertical displacement
and the corresponding postures of the relevant points, (b) is the vertical displacement trajectory of the
hip and the corresponding postures of the relevant points. In this paper, the maximum and minimum
values of vertical displacement of the foot joints are taken as the rhythm points. In this paper, the
maximum and minimum values of the vertical displacement of the foot joints are taken as the
rhythmic points, where the maximum value of the foot joints corresponds to the lifting action of the
foot, and the minimum value of the foot joints corresponds to the landing action of the foot, and some
of the minimum and maximum values are not the corresponding rhythmic points, and the up-and-
down range of the left foot during the dance is no more than 72cm (16.22cm). The left foot does not
move up and down more than 72cm (16.22cm~88.12cm), and the hips do not move up and down
more than 93cm (122.01cm~214.34cm) during the dance, reflecting the fact that the female
performers in the Flower Drum Lanterns emphasize on the elegance and lightness of their dance
movements, and the most typical movement is the Wind Swinging Willow, which is very soft and
charming.
(a) The redirect displacement trajectory of the left
foot
(b) The vertical displacement of the hips
Figure 3. The “flower drum lamp” dance posture changes
4.2 Musical Characteristics of Non-heritage Dance Fragments in Northern Anhui Province
After processing the time delay matrix and enhancing the noise reduction, the note repetition pattern
of the soundtrack of the Northern Anhui non-heritage dance can be extracted. 11.1 s of audio
segmentation, the segmented audio waveforms are shown in Fig. 4, which shows that the length of
the music segments ranges from 0.44 s to 2.04 s. The reason is that the melody of the music is not
strictly repetitive, which is also in line with the formal characteristics of the Northern Anhui non-
heritage dance. By selecting music segments with different rhythmic lengths, the music with non-
strictly repetitive melodies was segmented evenly, and the music was segmented at certain intervals
according to the beat lengths obtained from the rhythmic extraction analysis. 11.1 seconds of audio
was segmented, and the corresponding beat lengths yielded 12 music segments, with an average
length of 0.925 seconds.
Xiaona Li. Applied Mathematics and Nonlinear Sciences, 9(1) (2024) 1-16
12
Figure 4. The audio waveform after the section
5 Dissemination and Inheritance of Non-Heritage Dance in Northern Anhui Province
In the era of new media, the most effective means of traditional culture dissemination is social
networks. Therefore, this paper takes social networks as the object of exploring the dissemination
effect of the virtual display of dance on the non-heritage dance culture in northern Anhui Province,
specifically selects the mainstream W platform as the medium, and grabs the relevant data for analysis.
5.1 Distribution of non-heritage dance resources in Northern Anhui
The annual distribution of non-heritage dance resources in northern Anhui on the W social media
platform is shown in Table 2. From the quantitative point of view, the number of non-heritage dance
resources in northern Anhui shows an increasing trend year by year, especially from 2020 to 2021,
which is the most significant growth trend. It can be seen that, along with the increasing emphasis on
intangible culture and the background of attaching importance to soft power, the dissemination of
these non-heritage dance resources in northern Anhui is increasing. Under the background of “soft
power,” the dissemination of the non-heritage dances in northern Anhui is constantly increasing. From
the values of browsing, liking, stepping on, and commenting, the total browsing volume of these non-
heritage dances is 29970600 times, and they have received a total of 3770030 liking, 128779 stepping
on, and 462091 commenting, in which the number of liking is much larger than the number of
stepping on. Thus, it can be seen that most of the audiences recognize the non-heritage dances in
northern Anhui, which may be largely due to the high quality of the content of the cultural videos,
which better satisfy the aesthetic needs of the audiences.
Table 2. The annual distribution of the dance resources in northern Anhuis
Year
Quantity
Views
Thumb up
Point step
Comment quantity
2013
180
2923649
724553
14420
46140
2014
350
3915132
144175
11272
48948
2015
644
2551315
262672
18888
33523
2016
920
3339416
148006
16141
65579
2017
1073
4232903
304835
12551
58352
2018
1103
3994926
724323
20144
67017
2019
1228
1474458
612545
7024
48488
2020
1361
3961643
328571
12521
55130
2021
1524
3577139
520350
15818
38914
Total
8383
29970600
3770030
128779
462091
Cultural Inheritance and Artistic Construction of Non-heritage Dance in Northern Anhui in the Era of
Artificial Intelligence
13
However, although the number of non-heritage dance resources in northern Anhui increases year by
year, the number of views, likes, steps, and comments does not show the same growth trend. The
distribution relationship of non-heritage dance resources in northern Anhui is shown in Fig. 5. Among
them, the total number of video views, likes, and steps reached the maximum in 2018, which were
724323, 20,144, and 67,017, respectively. It can be seen that although the number of these types of
resources has increased, there has been no significant improvement in quality. The means of these
data are much smaller than the standard deviation and larger than the median, which indicates a high
degree of dispersion. There is no significant improvement in quality. The mean of these data is much
smaller than the standard deviation and larger than the median, which indicates that the degree of data
dispersion is high, the dissemination effect of the non-heritage dance resources in northern Anhui
varies greatly, and the number of videos with high dissemination effect is only a minority. From the
relationship between the number of views and the number of likes, stepping on and commenting on
the videos, the number of views and the number of likes, stepping on and commenting on the videos
are all positively proportional to the number of views and the number of likes, stepping on and
commenting on the videos. In terms of the relationship between the number of views and the number
of likes, pedals, and comments, the number of views of the non-heritage dance resources in northern
Anhui is proportional to the number of likes, pedals, and comments, with the correlation degree of
likes > comments > pedals.
Figure 5. The distribution of the dance resources in northern Anhui
5.2 The intervening role of virtual display technology
In this paper, we reconstructed the non-heritage dance resource of northern Anhui through a virtual
display method and put it on a social platform. We examined its dissemination effect after a certain
period of time. The dissemination of the non-heritage dance resource of northern Anhui under the
intervention of the virtual display technology is shown in Table 3. The table shows the duration of
time after the virtual dance resource was released. The reconstruction of the non-heritage dance
resource based on the virtual display method has increased the amount of the steady dissemination of
the resource by about 400 retweets, which is nearly twice as much as that of the original. The reason
why the forwarding scale of the non-heritage dance resource can increase dramatically in a certain
period of time is that the information has been reconstructed by the virtual display method, which
indirectly obtains more forwarding volume. Furthermore, the virtual display’s intervention in the
process of information dissemination decreases the predictability of the information diffusion range.
The visualization of non-heritage dances in northern Anhui Province is conducive to broadening the
space of dance expression, plays an important role in the diversity of dance expression and forms of
Xiaona Li. Applied Mathematics and Nonlinear Sciences, 9(1) (2024) 1-16
14
expression, enhances the attractiveness and infectiousness of the works, and plays an important role
in the dissemination and development of dance culture.
Table 3. The forwarding capacity of the dance resources in northern Anhui
Time/min
Primary path
Virtual reconstruction
Time/min
Primary path
Virtual reconstruction
50
206
208
1300
300
677
100
208
211
1350
301
684
150
209
216
1400
314
686
200
215
243
1450
317
690
250
217
250
1500
318
699
300
217
281
1550
319
702
350
225
306
1600
319
703
400
230
310
1650
328
709
450
247
323
1700
331
716
500
247
455
1750
332
723
550
252
524
1800
332
731
600
254
591
1850
333
734
650
257
621
1900
333
737
700
271
624
1950
335
743
750
271
628
2000
345
746
800
274
631
2050
345
748
850
276
644
2100
349
750
900
277
646
2150
362
755
950
277
651
2200
364
766
1000
282
652
2250
368
775
1050
289
658
2300
369
776
1100
291
669
2350
370
777
1150
296
670
2400
378
784
1200
296
673
2450
380
789
1250
299
674
2500
396
796
6 Conclusion
This paper explores the potential for innovative preservation and dissemination of traditional dance
in the artificial intelligence era. Through detailed analysis of dance videos, specifically the “Flower
Drum Lantern” dance, we have successfully extracted key movement features and rhythms. Our
findings reveal that the dance’s dynamic beauty is characterized by a maximum vertical displacement
of 72 cm for the feet and 93 cm for the hips. These metrics not only highlight the dance’s aesthetic
appeal but also offer a scientific foundation for its transmission and instruction.
Further, our study examines the impact of virtual display technology on social media platforms,
noting a significant enhancement in the dance’s dissemination efficiency. The adoption of virtual
presentation methods led to an approximately 400% increase in video retweets, underscoring the vast
Cultural Inheritance and Artistic Construction of Non-heritage Dance in Northern Anhui in the Era of
Artificial Intelligence
15
potential for spreading cultural heritage in the new media landscape. Audience engagement, as
measured by likes and comments, also reflects a positive reception to this innovative presentation
style.
By integrating artificial intelligence, motion capture, and virtual reality technologies, we have not
only preserved the essence of traditional dances but also expanded their reach and engaged a wider
audience in appreciating intangible cultural heritage (ICH). This research offers fresh perspectives
and methodologies for safeguarding ICH, providing valuable insights for its preservation across
different regions and forms.
Funding:
1. 2023 Anhui Province University Philosophy and Social Science Research Project: A Study on
the Ecological Changes of Huaibei Flower Drum Opera after Its Successful Application for World
Heritage under the Background of “Cultural Confidence” (2023AH052215).
2. 2022 Cebu College Teaching Research Project: Research on Teaching Reform and Educational
Quality Monitoring System of Dance Performance Major under the Perspective of Professional
Assessment (szxy2022jyxm43).
3. Cebu College Quality Project 2023: Cebu College School of Modern Industries for the
Integration of Arts and Performing Arts Industry and Education (szxy2023cyxy03).
References
[1] Zhu, W. S., & Liu, X. (2020). Vernacular landscapes building and protection of meme in the dong ethnic
minority regions. Social Science Quarterly, 101.
[2] Cerquetti, M., Nanni, C., & Vitale, C. (2019). Managing the landscape as a common good? evidence from
the case of “mutonia” (italy). Land Use Policy, 87.
[3] Yeh, H. R., Lin, L. Z., & Lu, C. F. (2019). Classification of traditional cultural elements in temple street
festivals using the fuzzy kano model. Current Issues in Tourism, 22(6-10), 1190-1215.
[4] Tamm, M. (2021). A lexandra s terling -h ellenbrand. medieval literature on display: heritage and culture
in modern germany. The American Historical Review.
[5] Soopramanien, R. (2017). International trade in indigenous cultural heritage: what protection does
international law provide for indigenous cultural goods and services in international commerce?. stanford
journal of international law, 53(2), 225-248.
[6] Singal, P., & Chopra, A. (2023). Cultural aspects of mental health in south asians. Psychiatric annals.
[7] Chen, Q., & Zhang, R. (2019). The power of confucian heritage culture. Asia Pacific Journal of Education.
[8] Li, X., Karuppiah, M., & Shanmugam, B. (2021). Psychological perceptual analysis based on dance
therapy using artificial intelligence techniques. International Journal on Artificial Intelligence Tools.
[9] Tang, Y., Zhou, L., Cao, J., Li, J., & Nie, X. (2018). Integration of digital cultural heritage resources in
china: understanding public expectations. Libri, 68(1), 59-70.
[10] Roszczynskakurasinska, M., Domaradzka, A., Wnuk, A., & Oleksy, T. (2021). Intrinsic value and
perceived essentialism of culture heritage sites as tools for planning interventions. Sustainability, 13(9),
5078.
[11] Carr, J., & Baddoo, D. (2020). Dance, diaspora and the role of the archives: a dialogic reflection upon the
black dance archives project (uk). Dance Research, 38(1), 65-81.
Xiaona Li. Applied Mathematics and Nonlinear Sciences, 9(1) (2024) 1-16
16
[12] Zhong, Y. X. (2017). Artificial intelligence: concept, approach and opportunity. Chinese Science Bulletin,
62(22), 2473-2479.
[13] Hu, M., & Wang, J. (2021). Artificial intelligence in dance education: dance for students with special
educational needs. Technology in Society, 67, 101784-.
[14] Peng, H., Hu, J., Wang, H., Ren, H., & Li, J. (2021). Multiple visual feature integration based automatic
aesthetics evaluation of robotic dance motions. Information (Switzerland), 12(3), 95.
[15] Ozcimder, K., Kong, Z., Wang, S., & Baillieul, J. (2018). Perceiving artistic expression: a formal
exploration of performance art salsa. IEEE Access, PP(99), 1-1.
[16] Zhuang, X., Yao, Y., & Li, J. (2019). Socio-cultural impacts of tourism on residents of world cultural
heritage sites in china. Sustainability, 11(3).
[17] Candela, G., Escobar, P., Carrasco, R. C., & Marco-Such, M. (2018). A linked open data framework to
enhance the discoverability and impact of culture heritage. Journal of Information Science.
[18] Karadeniz, C., & Ildir, Z. (2017). The views of prospective classroom teachers on cultural heritage-themed
museum education course. The Anthropologist.
