Available online at www.sciencedirect.

com

Procedia Engineering 41 (2012) 388 394

International Symposium on Robotics and Intelligent Sensors 2012 (IRIS 2012)

Measurement of the Flexible Bending Force of the Index and Middle

Fingers for Virtual Interaction
Nazrul H. Adnana,b,c, Khairunizam WANa,b, Shahriman A.B.a,b,
SK Zaabaa,b, Shafriza nisha, BASAHa,b, Zuradzman M. Razlana,b, D.Hazry b, M. Nasir Ayobb,
Rudzuan M.Nora,b and Azri A. AZIZa,b
a

Advanced Intelligent Computing and Sustainability UNIMAP Research Group

School Of Mechatronic Engineering, Universiti Malaysia Perlis, 02600 Arau, Malaysia
c
Institut Kemahiran MARA Beseri Perlis, Malaysia

Abstract
In this paper the development of a new low cost dataglove based on fingertip bending tracking techniques for measuring the fingers
bending on various virtual interaction activities is presented as an alternative to the rehabilitation services enhancement in the betterment
of the quality of life especially for the disabled person. The purpose of the research is to design a flexible control for measurement study
of virtual interaction of index and middle fingers that are important in a variety of contexts as well as the deterministic approach. These
analyses of fingers flexing of the system were using the flexible bend sensor functioning as a key intermediate of the process to track the
fingertip positions and orientations. The main propose of the low cost dataglove is to provide natural input control of interaction in virtual,
multimodal and tele-presence environments as an input devices provide as they can monitor the dexterity and flexibility characteristics of
the human hand motion. Preliminary experimental results have shown that the dataglove capable to measure several human Degree of
Freedom (DoF), translating them into commands for the interaction in the virtual world.

2012 The Authors. Published by Elsevier Ltd. Selection and/or peer-review under responsibility of the Centre of
Humanoid Robots and Bio-Sensor (HuRoBs), Faculty of Mechanical Engineering, Universiti Teknologi MARA.
Open access under CC BY-NC-ND license.

Keywords: dataglove; virtual interaction; rehabilitation services enhancement; grip force; flexible control

1. Introduction
Dataglove is one of the most commonly used technologies for the gestural interaction in the human-computer interface.
Dataglove helps users to interface with the virtual world, and serves as an input device in non-verbal human communication
[1]. A commercial datagloves usually use expensive motion-sensing fibers and motion analyzers, and are consequently too
costly for the consumer market [2]. Nowadays, there are various methods of using sensing technology applications in order
to improve the research and development for datagloves application. For the example, Fahn, C.-S et al [3] used the
datagloves that are able to provide the several high accuracy, high reliability, and high capability in measuring the DOF of
human hands. Most of them are constructed by using the scientific sensor devices in order to measure the finger bending
angles. Oz et al. [4] explained that by using the existing system of a sensory Cyberglove assisted with the ability of Flock of
Birds 3-D motion tracker, the resulting system has managed to translate 26 American Sign Language (ASL) alphabets on a
very impressive number of extracting gesture rate equally to 96.0% by using the artificial neural network (ANN) approach.
The development of the dataglove for measuring the flexible bending force uses of flex sensor called as Flexible Bend
Sensor as shown in Fig. 1 were embedded into a hand glove attached at the top surface of the dataglove. The basic operation
of the sensor is simple, when the finger is bending, at the same time sensor is bend and the DoF generated output result can
be measured. Some of measured output results are able to be calculated into a various application including grasping force.

1877-7058 2012 Published by Elsevier Ltd. Open access under CC BY-NC-ND license.
doi:10.1016/j.proeng.2012.07.189

Nazrul H. Adnan et al. / Procedia Engineering 41 (2012) 388 394

Nazrul et al. [5] stated that the development of the GloveMAP system based on the flexible bend sensor are able to
measure on the various force accordingly to the variety grasping techniques of the human or users hand. In order to
creating the ability to achieve high capacity in measuring the DOF of fingers, the number of sensors that available on each
research and development of the low cost dataglove leads to more use in virtual interaction of the virtual reality (VR) series.
The structure for this research paper starts with Section 2, which addresses about the literature review of the several
related researches for the example; study approaches, product applications and research problems of recognizing the
dataglove. For Section 3, the topics will be describing the methodologies of the system. Section 4 continues with the
experimental design method including the experimental setup and results. Finally, Section 5 will present the conclusions and
proposing some possible future work.

Fig. 1.Flexible Bend Sensor 4.5 [9]

2. Literature Reviews
Capacity and capability to realize the use of technology adopted from engineering-based applications has grown rapidly,
nowadays many researches and innovation products have been developed and produced. For example, Yu-Fen Fu et al. [6]
stated that the uses of sign language by using more modern methods typically combine a large-scale movement are much
easy to identify. The authors also said that the system produced is much capable to translate the sign language for
handicapped-aphasia person and involves in the small-scale of finger bending movement and its requiring a more subtle
recognition system [6]. Deborah et al. [7] explain that the main problem for the sign language is involves in a large-scale
actions, for example, if the activities contain the several gestures movement which is combining at all movements including
the finger, palm, arm, and others. At the same time it is quite difficult for those who cannot manage their activities alone
(handicapped person). In this case, the study on patients behavior shows that the finger performs only at small-scale
actions, e.g., the movement of finger actions. The required solution must be the same with the other methods that can help
severely disabled person to communicate at the basic level with others in the 'language of the finger'.
To detect the position and orientation of the fingertips, the use of two sensors and two coil module generators can be
installed at the end of the metacarpal finger [18-20]. The procedure output result of fingers orientation can be manipulated
into a virtual environment or tele-operation system [8-12]. Y. Han [13] reports on the application of the visual motion of
dataglove, which is composed by an inexpensive consumer glove with attached thin-bar-type optical indicators and a
closed-form reconstruction algorithm that can overcome the common disadvantages of single-channel video approaches,
which is occlusion problem and the study needs for inconvenient iterative reconstruction algorithms. According to the J.
Rao et al. [14] stated that the used of the k-mean algorithms as an alternative to produce the development of low cost
dataglove that able to detect the hand learning and recognition gesture system by using the hand gesture recognition
approach. The approach is applied to cluster and finally it can be justify at the several data from sensor by using the hidden
Markov (HMM) model.
Meanwhile for the other approaches, Shuai J. et al. [21] stated that the designed dataglove is to improve the accuracy of
data captured and the efficiency of the calibration method. The main propose of the system is to establish a mapping
between the actual angle of the target and sensor readings from the sensor relative to the cross-over effects of constraints
based on anatomical. There are many researcher were study about how to produced and developed for tracking of finger
bending and fingertip pressing. At the same time the dataglove input device capable to provide controls the natural input
interaction in virtual mode and tele-presence environment because they can monitor the agility and flexibility characteristics
of human hand motion. For the example the sensory gloves are much popular to record and analyzed human fingers
movements in healthy subjects [22-27] and patients [28].

389

390

Nazrul H. Adnan et al. / Procedia Engineering 41 (2012) 388 394

3. Methodologies
3.1 Mathematical model of human hand

Fig. 2. Hand's Kinematic Structure [3]

To realize the use of mathematical modeling in the study of the kinematic structure of the hand, there are some
significant part of the structure to measure the movement of the finger, it is divided into four Proximal Interphalangeal
(PIP) joints and a Interphalangeal (IP) joint in developing a low cost Dataglove as shown in Fig.2. The use of two sensors
on the index and middle finger in realizing the study on mathematical models is essential because it is needed to detect the
movement of a PIP or IP joint for both index and middle fingers [3]. The sensor on the surface of the finger is capable to
detect the movement of an Metacarpophalangeal (MCP) and be able to measuring more to 10 DOF of the hand. With a so
much variety of finger DOF, then the movement of the finger become highly constrained, so that it can not make gestures
indiscriminately. There are lot of the constraints motions examples, the DOF of fingers can not bend in the backwards
direction too much. The motion constraints that there are a capabilities to simplifying the research and the development of
the Dataglove can be typically classified as follows [3].
A. Intrafinger constraints: This is a common constraint on the movement of the same finger joints. Using this
constraint, all types of movement, particularly all fingers will be much easier to using the following equation:
= 0.67

(1)

Where
is the stated as the fingers bending angle of the DIP joints and
PIP joint bending angle.

is called as the structure for the

B. Angle range constraints: These types of difficulty constraint are refers to the limits of the ranges between the
finger motions of hand anatomy. It is usually within the following ranges:
0
0
0

(90 ~100 )

_
_

_
_

90
90

and,
(2)

The proposed dataglove is planned to be able to measure several DOF, and focuses on two fingers only, however,
studies and experiments are in progress. Mathematical formulation as shown in Eq. 1 and Eq. 2 could be used as a reference
to design the dataglove.
3.2 The construction of dataglove
A flex sensor is an analogue resistor, and it works as a variable analogue voltage divider as shown in Fig.3. The basic
constructions of the flex sensor consist of the carbon resistive elements, which is the operation of flex sensor is substrate
within a flexible thin element. Mean that the element will be less resistance if the carbon is more. The basic flex sensor
framework is when the element of thin is bent then the sensor will be produces a higher resistance output value and then
vice versa. With a typical flex sensor, by referring to the external information a flex of 0 degrees gives 10K resistance, and

391

Nazrul H. Adnan et al. / Procedia Engineering 41 (2012) 388 394

if the flex sensor is bending into 90 degrees, it will gives approximately 30-40K [14,15,16]. Based on these outputs, the
bending angles of the fingers are calculated.

(a)

(b)

Fig. 3. (a) Flex Sensor Voltage Divider, (b) Bend Sensor "Resistance to Voltage Converter [15]

4. Experiments
Experiment was carried out by using two flex sensors, and the purpose is to show the reliability of flex sensors on the
sign language translation via virtual image / interaction. By using the movement of the index finger and middle finger so the
resistance will be measured and evaluated by ensuring that the resulting signal can be analyzed. Then, the signal had to be
sent to the microcontroller called as Arduino Uno [17] until the last signal will then be evaluated and interpreted before
being sent to the self developed programme.
4.1. Experimental Setup.
For the experimental setup, the ability to interpret the dataglove for sign language into the virtual word language shows
the potentials adaptability of the disabled usage. Figure 4 shows the low cost dataglove which capable to measure index and
middle finger movement and translate up to 7 potentials virtual alphabet Word A, B, C, D, F, K and number 8. The result
shows that all the fingers static bending of the finger movement can be recorded. Secondly the process of flex-sensors
bending that attached to the hand-glove could be initiates the process. The most significant component in this design is the
flex sensor, because it requires negligible force when the sensor is bend. Two flexible bend sensors were embedded into
the dataglove, whereas each finger having a different functions. The signal from the flex sensor will be interact with
MatLAB software V- Realms Builder 2.0. Throughout the IF Rule, the movement of index and middle fingers were result
the output based on the fuzzy logic controller. Fuzzy logic controller is employed to be corresponding to the relation
between the movements of the fingers with the alphabet. The basic connection of the Flex Sensor system can be than
connected directly to the targeted system which is acquire data relative to the amount of bending that occurs.

Sensor #1(Index)

Sensor #2 (Middle)

Fig. 4. Bending activities (a) straighten fingers (b) bending of index finger (c) bending of middle finger (d) bending of both fingers [17]

392

Nazrul H. Adnan et al. / Procedia Engineering 41 (2012) 388 394

Virtual Alphabet

Fig. 5. Interaction with virtual interaction

Figure 5 shows the experiment of how the appeared alphabet comes from the movement of index and middle fingers
gestures. Alphabet F for the example may not a characterised by its value on the index finger alone but instead it may have
middle finger also. In this situation, its deal with two variables fingers index and middle. The joint probability mass
functions of the two variables may be written as below whereas P: Joint Probability.
P (Index =

, Middle =

)=

(3)

,
and
,
then each
If the variables integer values for Index and Middle fingers takes in the range of
combination of Index and Middle values is equally likely, the joint probability density of Index and Middle can be
formulated as below and the formulation shown m: Index, n: Middle fingers.

=1

(4)

4.2. Experimental Results.

For the experimental results, several activities of the fingers motion have been taken. Starting with a straight finger
gestures until the fingers to bend by 45 degrees, finger bending process is intended to provide an appropriate signal to the
microprocessor then the signal from the Arduino transmitting directly to the virtual program that can make it translating the
sign language words. Figure 5 above shows the activities undertaken during the interpretation of the movement of the index
finger and middle finger gestures as well as the interaction of the movements of index and middle with the virtual alphabet
stored in the computer.
Decisions resulting from these experiments have already been taken and analyzed. Figure 6 below shows that the
suitability of the sensor in convenient formulation of consumerism and consistency with the objectives stated earlier, bend
with a more scientific method. Progress data is taken into account by 100 data for each movement and each data will be
evaluated by the rate of voltage change. Thus, each value of voltage changes resulting from the bent finger will then be sent
to the motor control circuit which is at the next level.
Figure 6a shows the relationship between the voltage versus time is clarify when there is no signal was produced during
the two flex sensors in situations where there is no bending activities occurs at both the index and middle fingers. The
situation will be changed when a Sensor #1 that is located at the index finger is bending, and then the waveform has
changed with the maximum reading as shown in Fig. 6b. However, Sensor #2 is still remains unchanged because there is no
reading or no wave is generated. In Fig. 6c, the situations are totally changed into different contrary to the output method.

Nazrul H. Adnan et al. / Procedia Engineering 41 (2012) 388 394

Waveform on Sensor #1 remain in initial position compare to Sensor #2 whereas the signal / waveform at a maximum
values. Finally for the Fig. 6d, both signal / waveform for Sensor #1 and #2 are totally at their maximum values of voltage.
The situation happened when both index and middle fingers are bending.

Fig. 6. Finger activities (a) straighten fingers (b) bending of index finger (c) bending of middle finger (d) bending of both fingers

5. Conclusions
In this paper, a new development of a low cost dataglove by using the flexible bend sensor which is able to recognize the
human fingers activities is presented. The goal of this research is to produce a product that is able to assist the disabled
person to perform their daily activities with no restrictions. Besides that, the proposed dataglove could also be applied to all
over the government and private hospitals in the sense to facilitate the services provided to patients and hospitals staffs.
The construction and production application of this system is using MatLAB software in associate with the Fuzzy Logic
translation system which is driven to be the perfect development of low cost dataglove. The production or translation result
is done only on 6 words A, B, C, D, F, and K and also the number of 8, however this research is only focused on two
main fingers index finger and middle finger in order to shows the potentiality and continuity for both fingers with the
developed system. For the future research, the study will be emphasized into the other fingers such as thumb, ring and little
finger. Therefore, the planning for improving the design and development for these gloves will be more specialized on data
or to be more focused on the other words translation for the example others 26 alphabet and 10 numbers in the basic sign
language communication. To this end, the proposed a low cost dataglove will be absorbed into the medical sector
applications as well as the sectors that need a virtual reality technology in term of the terminology of their application.
Acknowledgments
Special thanks to all members of UNIMAP Advanced Intelligent Computing and Sustainability Research Group and
School Of Mechatronics Engineering, Universiti Malaysia Perlis (UNIMAP) for providing the research equipments and
internal foundations. This work is supported by the Prototype Research Grant Scheme by the Ministry of Higher Education
Malaysia to Universiti Malaysia Perlis (PRGS 9003-00313).
References
[1]
[2]
[3]

M. Grimm, A. A. Arroyo, and M. C. Nechyba, Thing, 2002 A robotic hand with realistic thumb pronation Florida Conference on Recent Advances
in Robotics, Miami
B. Takacs., 2008. How and why affordable virtual reality shapes the future of education, The International Journal of Virtual Reality, vol. 7, no. 1, pp.
53-66.
C. S. Fahn and H. Sun, 2005. Development of a Dataglove With Reducing Sensors Based on Magnetic Induction, IEEE Transactions On Industrial
Electronics, Vol. 52, No. 2.

393

394
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]

[19]
[20]
[21]
[22]
[23]
[24]
[25]
[26]
[27]
[28]

Nazrul H. Adnan et al. / Procedia Engineering 41 (2012) 388 394

Oz, C., Leu, M.C., 2005. Recognition of Finger Spelling of American Sign Language with Artificial Neural Network Using Position/Orientation
Sensors and Dataglove, ISNN.
Nazrul H Adnan, Khairunizam WAN and Shahriman AB, 2012. Accurate Measurement of the Force Sensor for Intermediate and Proximal Phalanges
of Index Finger, International Journal of Computer Applications 45(15):59-65.
Yu-Fen Fu and Cheng-Seen Ho, 2007. Static Finger Language Recognition for Handicapped Aphasiacs, IEEE
Deborah Bessonny Gilden, and Brad Smallridge, 1993 Touching Reality: A Robotic Finger spelling Hand for Deaf-Blind Persons, Proceedings of
Virtual Reality Conference
Sing, B.K.; Ikeuchi, K., 1995 A robot system that observes and replicates grasping tasks, In Proceedings of 5th International Conference Computer
Vision, Cambridge, MA, USA; pp. 1093-1099.
Toney, A., 1998 A novel method for joint motion sensing on a wearable computer, In Proceedings of 2nd International Symposium Wearable
Computer, Pittsburgh, PA, USA; pp. 158-159.
Perng, J.K.; Fisher, B.; Hollar, S.; Pister, K.S.J., 1999 Acceleration sensing glove (ASG), In Proceedings of 3rd Intnational Symposium Wearable
Computer, San Francisco, CA, USA; pp. 178-180.
Dogramadzi, S.; Allen, C.R.; Bell, G.D.; Rowland, R., 1999 An electromagnetic imaging system for remote sign language communication, In
Proceedings of 16th IEEE International Conference Information and Measurement Technology, Venice, Italy,; pp. 1443-1446.
Tarchanidis, K.N.; Lygouras, J.N., 2001 Dataglove with a force sensor, In Proceedings of 18th IEEE Conference Information and Measurement
Technology, Budapest, Hungary; pp. 380-385.
Youngmo Han, 2010. A Low-Cost Visual Motion Dataglove as an Input Device to Interpret Human Hand Gestures, IEEE Transactions on
Consumer Electronics, Vol. 56, No. 2.
Jinjun Rao, Tongyue Gao, Zhenbang Gong and Zhen Jiang, 2011 Low Cost Hand Gesture Learning and Recognition System Based on Hidden
Markov Model, Second International Symposium on Information Science and Engineering, IEEE, 433-438
Information on http://www.myduino.com/index.php?route=product/category&path=39
Information on http://www.cytron.com.my/usr_attachment/SN-FLX-01.pdf
Information on http://arduino.cc/en/Main/arduinoBoardUno
Nazrul Hamizi Adnan, Wan Khairunizam Wan Ahmad, Shahriman Abu Bakar, Siti Khadijah Zaba, Hazry Desa and Mohd Azri Abd Aziz, 2012
The Development of a Low Cost Dataglove by Using Flexible Bend Sensor for Resistive Interfaces, In Proceedings of 2nd International MalaysiaIreland Joint Symposium on Engineering, Science and Business (IMiEJS).
Chin-Shyurng Fahn, and Herman Sun, 2010. Development of a Fingertip Glove Equipped with Magnetic Tracking Sensors, Sensors, 10, 1119-1140
Nazrul Hamizi Adnan and Wan Khairunizam Wan Ahmad, 2012 Gesture Recognition Based On Hand Postures And Trajectories By Using
Dataglove: A FuzzyProbability Approach A Review, In Proceedings of International Conference on Man-Machine Systems (ICoMMS).
Shuai Jin, Yi Li and Weidong Chen, 2010 A Novel Dataglove Calibration Method, The 5th International Conference on Computer Science &
Education Hefei, China. August 2427.
Boesnach I, Moldenhauer J, Burgmer C, Beth T, Wank V, Bs K, 2004 Classification of phases in human motions by neural networks and hidden
Markov models, In: IEEE international conference on cybernetics and intelligent systems. PP.97681.
Jerde TE, Soechting JF, Flanders M., 2003. Biological constraints simplify the recognition of hand shapes, IEEE Trans Biomed Eng. 50:2659.
Lang CE, Schieber MH., 2003. Differential impairment of individuated finger movements in humans after damage to the motor cortex or the
corticospinal tract, J Neurophysiol. 90:116070.
Mason CR, Gomez JE, Ebner TJ. , 2001, Hand synergies during reach-to-grasp. J Neurophysiol. 86:2896910.
Santello M, Flanders M, Soechting JF, 1998. Postural hand synergies for tool use. J Neurosci. 18:1010515.
Weiss EJ, Flanders M., 2004. Muscular and postural synergies of the human hand. J Neurophysiol.92:52335.
Lang CE, Schieber MH., 2004, Reduced muscle selectivity during individuated finger movements in humans after damage to the motor cortex or
corticospinal tract. J Neurophysiol;91:172233.