The Correlation of Lux and Distance to Markerless Augmented

Reality Detection Technique for Digital Therapy Application

Maulidya Anggraini

, Akuwan Saleh

, Hani’ah Mahmudah

, I Gede Puja Astawa,

Aries Pratiarso and Muhammad Zen Samsono Hadi

Department of Electrical Engineering, Politeknik Elektronika Negeri Surabaya, Sukolilo, Surabaya, Indonesia

zenhadi@pens.ac.id

Keywords: Correlation, Augmented Reality, Markerless, Digital Therapy Application.

Abstract: Considering the huge number of autistic children in Indonesia, it appears that many parents of autistic

children do not have a good understanding of how to specifically treat autistic children. A consultation with

a psychologist for therapy cannot be undertaken at any time or place, according to a treatment perspective.

Digital therapy applications are reportedly being considered as a solution to these problems in the current

digital era. Markerless augmented reality has the potential to be one of the innovative technologies with

interactive features that can create a visual world that resembles the real world. There are several important

aspects associated with measuring the accuracy of the tracking surface and object appearance. The purpose

of this paper is to determine the connection between lux and distance to build markerless augmented reality

capable of detecting long distances with minimal lux. The proposed method was tested at different

locations, lux, and distances in the range 2 to 200 cm. The result, it can connect the design between distance

and lux to show therapy objects in both indoor and outdoor areas using a markerless augmented reality

method.

1 INTRODUCTION

Noticeably, there is a persistent rising number of

autistic children per year. The World Health

Organization (WHO) stated that autistic children's

rate experiences a remarkable rise year after year.

The prevalence range with regards to the number of

autistic children in the local context is 1.14 per 1000

or 1 in 87 children (Alshaban et al., 2019), while the

prevalence range in the United States is 16.8 per

1000 children. This statistical number could undergo

a rise if there is an attempt to screen every child

across the country. Autism is a sort of

developmental disorder in a child whose symptoms

have already appeared before the child reaches the

age of three years. It occurs due to a severe

neurobiological disorder that affects brain function

hence the child is incapable to interact and

communicate effectively with the outside world.

Associate with this issue, parents are demanded to

have such a good understanding regarding certain

things around autism and be capable to organize

therapy exercises for their children. Autistic children

undergo such difficulties in terms of recognizing and

expressing emotions as well. Most educators and

psychologists are in the same agreement that their

emotions can affect their ability to focus on a task.

In addition to having several characters such as

hyperactive or unable to remain silent, most autistic

children perform repetition tasks due to the boredom

and lack of attractiveness while encountering certain

objects. This occur sowing to the presence of a

disorder in the child's brain function called Attention

Deficit Hyperactivity Disorder (ADHD). As a

consequence, autistic children frequently spend a

huge time resting and undergo such difficulties in

terms of retaining their focus (Escobedo et al.,

2014).

Accordingly, innovation is highly needed by

utilizing digital technology to provide therapeutic

treatment for autistic children. Some technologies

that previously have been applied as therapeutic

media comprise digital libraries in the form of

https://orcid.org/0000-0002-5639-3064

https://orcid.org/0000-0002-9082-1448

htt

s://orcid.or

/0000-0002-1675-2077

956

Anggraini, M., Saleh, A., Mahmudah, H., Astawa, I., Pratiarso, A. and Hadi, M.

The Correlation of Lux and Distance to Markerless Augmented Reality Detection Technique for Digital Therapy Application.

DOI: 10.5220/0010957200003260

In Proceedings of the 4th International Conference on Applied Science and Technology on Engineering Science (iCAST-ES 2021), pages 956-962

ISBN: 978-989-758-615-6; ISSN: 2975-8246

 2023 by SCITEPRESS – Science and Technology Publications, Lda. Under CC license (CC BY-NC-ND 4.0)

pocket pc consisting of image-based communication

media and user data storage (Leroy et al., 2005),

emotion-based video games with artificial

intelligence (Irani et al., 2018), speech therapy for

autism monitoring mobile application (Santiputri et

al., n.d.), and audio recording-based therapy

applications to advance the quality of data records.

Nevertheless, some of these applications remain able

to obtain such development again with more

interactive technology to complete the existing

systems.

One of the alternative technologies that can be

exploited is the augmented reality. Augmented

reality is a current technology that enables the

introduction of elements both objects and computer-

generated locations based on the real world view.

There are two methods used by augmented reality,

namely marker detection and markerless detection.

The marker detection method has been applied to

find out the short length of a marker that can be

detected to execute on an embedded system with

low computing requirements. This research

conducted on indoor and outdoor environmental

conditions has three scenarios to research the

number of markers detected by the system. The

augmented reality also has the advantage of a quick,

straightforward, and robust process under changing

lux and distance conditions. On top of that, the level

of accuracy and throughput in terms of detecting

markers and acquiring results in different

environments possess a fairly high level (Díaz et al.,

2018). The usage of augmented reality in the form of

applications has also been administered to health

care, sports, military, security systems (Díaz et al.,

2018), and virtual laboratory applications (Abhishek

et al., 2019).

Nevertheless, back to the major target issue of

this paper discussed is those who are autistic

children with the hyperactive character or unable to

be silent, and a more precise method to use is

markerless augmented reality. Markerless

augmented reality possesses the ability to detect a

certain point in a quick time by using one of the

algorithms in ARCore namely environmental

understanding. With this algorithm, ARCore will

explore clusters of feature points that appear to take

place on general horizontal or vertical surfaces, such

as tables, walls, and other textured surfaces. In other

research (Singh et al., 2018), it is also demonstrated

that augmented reality can be utilized for accurate

measurements based on focal distance, age, and

brightness within the distance limit of 33.3 to 50 cm.

So that both marker detection and markerless

detection seem to demand a deeper paper to detect

objects based on the parameters of lux and distance

to the environment. The limitation of this paper is

that it can only detect objects from a distance of 50

cm. The proposed markerless augmented reality

method in this paper has a distance more than 50 cm.

Besides that, this method has the advantage of being

able to detect movement, which is relevant for

autistic children who have trouble focusing. In other

papers, determining the correlation value is also

necessary to understand the connection between the

two parameters, namely the distance value and the

lux value (Astuti, 2017).

This paper proposed the correlation of distance

and lux to demonstrate objects video tutorials as a

therapeutic media for autistic children using

markerless augmented reality. The calculations

presented in the test section include the average

value, median, standard deviation, and covariance

for estimating the size of the data center. Ultimately,

this paper aims to measure the accuracy of the

system known through the results of its correlation

value. Furthermore, the influence for autistic

children themselves while using this application can

be noticed in an interaction between the application

and the location of scanning objects thus as to train

the sensory and motor skills of autistic children in

which the success of therapy also depends on the

severity of the symptoms while being reviewed from

the age factor of starting therapy and parental

support (Asrizal, 2016). The result, this technique

can be used to support digital therapy for autistic

children.

2 SYSTEM OVERVIEW

This section explores designing a system for

applying lux values and distance values in searching

markerless augmented reality-based correlation

values. Figure 1 demonstrates a diagram block of the

proposed research method. From the diagram block,

the discussion will be consisting of several points

ranging from markerless augmented reality design,

MAXST AR framework implementation, system

implementation, retrieval of lux value and distance

value data, and the size of the data centralization.

The system design is conducted from android

applications made using Unity software with C#

programming language as a method for designing

digital therapy applications for autistic children.

The Correlation of Lux and Distance to Markerless Augmented Reality Detection Technique for Digital Therapy Application

957

Figure 1: Research Diagram Block.

2.1 Markerless Augmented Reality

Design

The design of markerless augmented reality is

typically conducted in some stages. The first stage is

to invent scene tracking through Unity software

using Maxst AR. This paper will apply Instant

Tracker which is a technology to fine-tune the issued

object based on the field found in the camera image.

After the MAXST AR SDK is attached to Unity,

a single special scene is created for tracking in

which there is one canvas and button to start and

complete the tracking process. As a parameter of the

appearance of markerless objects, parameters are

given based on the Distance Sensor and Lux Sensor

that will simultaneously appear when the location

point tracking process begins. The scene tracking

picture is as shown in the Figure 2.

Figure 2: Object Tracking Implementation View of

Markerless Augmented Reality.

It can be noticed from the picture that there are

several elements in the object tracking scene

comprising the button for start and stop tracking, the

back button, and two sensors to measure the values

of both lux and distance programmed using the

language C#. With the relevant lux value and

distance value, markerless AR objects in the therapy

video for autistic children will be displayed. The

therapy video will display after tracking starts and

the video will still be at the first point of tracking.

When the camera is moved in another direction

beside the first tracking point, the camera will

present a real-world view and when it returns to the

first tracking point, the user can review the therapy

video as an object of markerless augmented reality.

Figure 3 presents the flowchart of the course of the

system in this research.

Figure 3: Proposed System Workflow.

2.2 Data Retrieval Process

Noticeably, there is a scenario in the process of

retrieving data in testing the markerless augmented

reality design. The data retrieval scenario owns two

parameters there are lux value and distance value.

The scenario aims to know the value of lux and the

value of optimal distance for tracking objects.

Moreover, it can also identify whether or not an

augmented reality markerless object appears at that

point. The process of data retrieval scenarios is

demonstrated as shown in Figure 4.

Figure 4: Data Retrieval Scenarios.

iCAST-ES 2021 - International Conference on Applied Science and Technology on Engineering Science

958

Data retrieval steps for markerless augmented

reality testing are as shown clearly from the diagram

block. The first step is the devise of markerless

design with Unity in which there is a design to load

scene tracking through Unity software using the

MAXST AR framework. Once the application is

successfully created in the smartphone, the tracking

point markerless augmented reality is then

determined. On this occasion, 10 points will be

tested in two-area conditions, namely in the indoor

and the outdoor area. Once the tracking point is set

up, the user can open this application through their

smartphone and do tracking at the specified location.

Once the application is open, the user can identify

and read both the lux and distance value at that

point. Furthermore, the user can start tracking by

pressing the start button according to the therapy

video to display. Moreover, the user can identify the

appearance of objects ranging from a distance of 2

cm to 200 cm with multiples of 2 cm. This is

conducted 10 times at the specified points and read

the lux value and write it down in the data retrieval

table.

Testing is managed under two environmental

conditions. In collecting data in the indoor area, it is

conducted at 07.00 – 10.30 (Western Indonesian

Time) in several indoor areas such as Kitchen,

Living Room, and Room Corners. This tracking is

undertaken in 10 points such as on prayer mats,

cardboard, tables, and rooms. Similar to data

collection in the indoor area, data collection in the

outdoor area is managed from06.00 –09.30(Western

Indonesian Time). A total of 280 data are taken with

a range of multiples of 2 ranging from 2 - 200 cm. In

this condition, there are also 10 tracking points such

as Courtyard, Terrace Plants, and several different

alleys.

2.3 Statistical Analytics

When conducting data testing using the results of

data collection in both indoor and outdoor areas,

statistical analytical calculations such as calculating

the average, median, and covariance values are

performed. The statistical data is calculated

immediately following the data collection process.

In this paper, two parameters are used: the distance

value and the lux value. The goal of this research

technique is to find the correlation value between the

distance value and the lux value. The analysis of the

obtained correlation values demonstrates the

system's accuracy. From the results of the data

collection process, the distribution of plot data in

both indoor and outdoor areas is shown in Figures 5

and 6.

Figure 5: Lux Data Distribution and Distance in Indoor

Area.

Figure 6: Lux Data Distribution and Distance in Outdoor

Area.

From Figure 5, it can be noticed that testing in

indoor area leads to the results stating that the

average appearance of Lux value at a distance of 2-

200 cm frequently occurs when lux is 0-30 lux so

that the approximate average value of lux is in the

range of 1-30 lux. Meanwhile, for outdoor testing in

Figure 6, it is demonstrated that the average

appearance of Lux value at a distance of 2-200 cm

frequently occurs when lux is 0-1500 lux. Therefore,

The Correlation of Lux and Distance to Markerless Augmented Reality Detection Technique for Digital Therapy Application

959







∑









the approximate average value of lux is in the range

of 0-1500 lux.

From the data of the acquisition of both distance

and lux value, it is later on processed to set up

statistical data. The first calculation is to find out the

average value of the lux through in (1)

 =

∑









(1)



By using the equation, it can be noticed that fi is

the frequency of the “i data group” and xi is the

middle value of the i-group of data thus the average

value of lux in indoor area obtained a value of 18.72

lux and the average lux in the outdoor area obtained

a value of 1240.4 lux, and then continued by looking

for the median value of lux by use in (2)

 = 

+ (



−







) (2)

The equation indicates that the variable  is the

bottom edge of the median class, 



is the

cumulative frequency, and  is the frequency of the

median class. After the calculation, the median value

of lux in indoor area obtained a value of 18.16 lux

and the median value of lux in the outdoor area

obtained a value of 1199.5 lux. After that,

calculations the standard deviation value in group

data and to find out this value can be obtained by use

in (3)

  =

√



−

∑



( − )

(3)

On the equation,  is the mean value and N is the

amount of data. By using (3), the standard deviation

value in the indoor area obtained a value of 8.67271

while the standard value of deviation in the outdoor

area obtained a value of 334,839. After gaining the

value of data centering size, it is supposed to look

for a summary of its measurements encompassing

covariance and correlation to look for a so-called

covariance value that is a measure of the combined

variability of two random variables. Afterward, it is

then supposed to perform calculation operations like

a (4)

 (, ) =

∑(

−

)(

−

)

(4)

From the calculations taken, it is obtained that

the value of covariance for the indoor area is -

7999.71 and for the outdoor area is -307165.82.

After obtaining the covariance value, the calculation

to determine the correlation value is then conducted

to know the tightness of the connection of two

variables in this case namely the lux value and

distance value. Regarding finding out the correlation

value, it is demanded first to know the value of

distance and its average and the value of lux and its

average. Once the required value is acquired, the

calculation can be continued use in (5)







(





)

−



(



)



(



)

(5)

√





−



(



)

√





−



(



)

Based on the equation, calculations are then carried

out and the correlation value in the indoor area

reaches out -0.52value while the outdoor area

reaches out -0.51 value. Due to having a negative

value, the connection between the two variables has

opposite characters in which the increase in distance

value will be accompanied by a decrease in the value

of the lux value.

3 EXPERIMENT AND

EVALUATION

The tests conducted in this paper are carried out in

the indoor area and the outdoor area ranging from a

distance of 2 cm to 200 cm with a range of multiples

of2.Thus,it is known that there are 280data obtained

from the results of data collection this time. In this

section, the results of the research will be displayed

to find out the best correlation value between the

distance value and the lux value. Previously, to

determine the actual average value of lux, it is

necessary to know each average value of lux per

centimeter (cm) as shown in both Figure 7 and

Figure 8 it can demonstrated that has two parameters

are lux value and distance value.

Figure 7: The Graph of Average Lux in Indoor Area.

iCAST-ES 2021 - International Conference on Applied Science and Technology on Engineering Science

960

Figure 8: The Graph of Average Lux in Outdoor Area.

According to Figure 7, the minimum distance for

data collection in the Indoor Room is 2 cm.

However, because all tracking objects at 10 points

cannot be displayed at a distance of 2 cm, the value

is set to 0. As a result, the graph demonstrates the

connection between the average lux value and the

distance value for the appearance of markerless

augmented reality objects. The figure demonstrates

that at a distance of 4-200 cm, the average lux that

covers each distance is in the 10-30 lux range.

Meanwhile, Figure 8 means the data was

collected at a distance of 2-200 cm, with the average

lux obtained covering each distance, namely in the

range of 0-1500 lux. It has a higher value at a

distance of 6 cm in the indoor space and 4 cm in the

outdoor space. This is due to the environmental

conditions, as data was collected at 07.00 WIB with

the camera facing direct sunlight, resulting in

comparatively high lux values.

As shown in Table 1, equation (5) is used to

calculate the correlation value. When the correlation

value obtained both indoor and outdoor is negative.

The number of data points, standard deviation, and

covariance as a reference parameter for the

correlation value were also included in the table.

Table 1: Table of Correlation Value Testing Results in

Indoor and Outdoor Area.

Location

Amount

of Data

Standart

Deviation

Covariant

Correlation

Value

Indoor

Area

280 8,67271 -7999,71 -0,52

Outdoor

Area

280 334,839 -307165,82 -0.51

In the table, a negative value indicates that the

correlation between the two variables is inverse.

When the distance between two points increases, the

value of lux decreases. This is coherent with the

graph, which shows that at distanceslessthan100cm,

the lux value obtained is relatively high. Meanwhile,

if the distance is slightly higher than 100 cm, the lux

value obtained is relatively low. To display video as

an object of digital therapy for autistic children,

application users should be in an area with proper

lighting and at a reasonable distance. To display

video as an object of digital therapy for autistic

children, application users should be in an area with

adequate lighting and at a reasonable distance. The

calculation of the correlation between lux values and

distance values in this paper is useful for

determining the recommended value for displaying

therapeutic video objects in the design of digital

therapy applications that children with autism can

use.

4 CONCLUSIONS

Based on the research, it is completely obvious that

this research utilizes augmented reality technology,

which enables the integration of elements, both

objects and computer-generated locations based on

real-world views. This technology is appropriate to

use to increase interest in digital therapy for children

with autism by using one of the methods, namely

markerless, because augmented reality also has the

advantage of a fast, straightforward, and strong

process in different lux and distance conditions.

This paper also includes calculations for determining

the correlation value between the lux value and the

distance value. Both obtain negative results, namely

-0.52 in the indoor area and -0.51 in the outdoor

area. The correlation value with the markerless

method can have an effect on the object's

appearance. When the markerless design detects

objects at a close distance, the lux value will be

high. The result, when using this application, users

should be in a well-lit area and at a reasonable

distance from the object tracking surface. According

to the results of tests with a minimum data collection

distance of 2 cm, video objects can only be

displayed in outdoor area. This happens because the

lux value, which is 60 lux, is sufficient.

Furthermore, the markerless design used has a

maximum distance of 200 cm. In other situations, if

the distance is greater than 200 cm, the video object

displayed as a therapy media will appear small and

difficult to see. For the future research, this

The Correlation of Lux and Distance to Markerless Augmented Reality Detection Technique for Digital Therapy Application

961

application can be added to a database that aims to

store data from the value of lux and distance in

indoor and outdoor area.

REFERENCES

Abhishek, M. T., Aswin, P. S., Akhil, N. C., Souban, A.,

Muhammedali, S. K., & Vial, A. (2019). Virtual Lab

Using Markerless Augmented Reality. Proceedings of

2018 IEEE International Conference on Teaching,

Assessment, and Learning for Engineering, TALE

2018, December, 1150–1153.

Alshaban, F., Aldosari, M., Al-Shammari, H., El-Hag, S.,

Ghazal, I., Tolefat, M., Ali, M., Kamal, M., Abdel

Aati, N., Abeidah, M., Saad, A. H., Dekair, L., Al

Khasawneh, M., Ramsay, K., & Fombonne, E. (2019).

Prevalence and correlates of autism spectrum

disorder in Qatar: a national study. Journal of Child

Psychology and Psychiatry and Allied Disciplines,

60(12), 1254–1268.

Asrizal. (2016). Penanganan Anak Autis dalam Interaksi

Sosial. Jurnal PKS, 15(1), 1–8.

Astuti, C. C. (2017). Analisis Korelasi untuk Mengetahui

Keeratan Hubungan antara Keaktifan Mahasiswa

dengan Hasil Belajar Akhir. Journal of Information

and Computer Technology Education, 1(April), 1–7.

Díaz, Á., Peña, D., & Villar, E. (2018). Short and long

distance marker detection technique in outdoor and

indoor environments for embedded systems. 2017

32nd Conference on Design of Circuits and Integrated

Systems, DCIS 2017 - Proceedings, 2017-Novem, 1–6.

Escobedo, L., Tentori, M., Quintana, E., Favela, J., &

Garcia-Rosas, D. (2014). Using augmented reality to

help children with autism stay focused. IEEE

Pervasive Computing, 13(1), 38–46.

Irani, A., Moradi, H., & Vahid, L. K. (2018). Autism

Screening Using a Video Game Based on Emotions.

2018 2nd National and 1st International Digital

Games Research Conference: Trends, Technologies,

and Applications, DGRC 2018, 40–45.

Leroy, G., Chuang, S., Huang, J., & Charlop-Christy, M.

H. (2005). Digital libraries on handhelds for autistic

children. Proceedings of the ACM/IEEE Joint

Conference on Digital Libraries, 387.

Santiputri, M., Sembiring, E. B., Janah, N. Z., & Mufidah,

M. K. (n.d.). Abang Manis : Speech Therapy for

Autism Monitoring Mobile Application. 3–8.

Singh, G., Ellis, S. R., & Swan, J. E. (2018). The Effect of

Focal Distance, Age, and Brightness on Near-Field

Augmented Reality Depth Matching. IEEE

Transactions on Visualization and Computer

Graphics, 2626 (c), 1–14.

iCAST-ES 2021 - International Conference on Applied Science and Technology on Engineering Science

962