New Directions for the Skincare using NFC based Battery-Free Approach with Deep Learning Techniques
- Alternative Title
- 딥러닝 기법에 의한 NFC 기반 배터리 프리 접근 방식의 새로운 스킨 케어
- Abstract
- Skincare has always been given a supreme importance in the field of dermatological science. In this thesis, a concrete discussion is carried on to support the skincare by developing a device using a completely new approach. A smart skincare device was developed that can harvest sufficient energy from the near field communication(NFC) based smartphone, thus elucidating the concept of battery-free design approach i.e. without requiring any battery cell to operate. The device mainly consists of two integrated sensors: One is skin moisture and the other is ultraviolet(UV) sensor. By conducting experimental tests on vivo with 8 subjects in indoor environment and 6 subjects in outdoor environment, skin moisture and skin temperature from the skin, ultraviolet A (UVA) and ultraviolet B (UVB) radiations from the sun were investigated. Later, six channel sensing output from the sensors including ambient humidity and ambient temperature were passed to the input of deep learning artificial neural network (ANN) model in order to predict the corresponding outputs and find the respective mean squared error(MSE). Classification of the ultraviolet index (UVI) outputs was done using the same ANN model by classifying into less harmful, moderate harmful and burn, therefore, the overall classification accuracy was found to be 99.8% which is in fact the best performance achieved by using ANN model. Moreover, it can also be noted that our skincare device is enclosed in a 3D flexible design print and is smart, battery-free with android application interface and lastly, it is convenient for anyone to take it everywhere at ease as compared to the other commercially available battery based devices.
피부 과학 분야에서 스킨 케어는 가장 중요하다. 이 논문에서는 완전히 새로운 접근법을 통해 장치를 개발하여 피부 관리를 지원하기 위한 구체적인 논의가 진행되었다. NFC(Near Field Communication) 기반 스마트 폰에서 충분한 에너지를 수확 할 수 있는 스마트 스킨 케어 장치가 개발되어 배터리가 필요 없는 설계 방식의 개념을 설명한다. 이 장치는 주로 두 개의 통합 센서로 구성된다. 하나는 피부 수분이고 다른 하나는 자외선(UV)센서이다. 실내 환경에서 8명의 피험자 및 실외 환경에서 6 명의 피험자로 생체에 대한 실험 테스트를 수행함으로써 피부로부터의 피부 수분 및 피부 온도, 태양으로부터의 자외선 A (UVA) 및 자외선 B(UVB) 를 조사 하였다. 추후 주변 습도 및 온도를 포함하는 센서로부터 6개의 채널 데이터가 해당 출력을 예측하고 각각의 평균 제곱 오차(MSE)를 찾기 위해 딥러닝 인공 신경망 모델(ANN)의 입력으로 전달되었다. 자외선 지수(UVI) 출력의 분류는 세단계로 분류하여 동일한 ANN 모델을 사용하여 수행되었으므로 전체 분류 정확도는 99.8%인 것으로 나타났다. 이는 실제로 ANN을 사용하여 달성 한 최고의 성능 모델 이다. 또한 우리의 스킨 케어 장치는 3D 플렉시블 디자인 프린트에 부착되어 있으며 안드로이드 응용 프로그램 인터페이스를 통해 동작하고 배터리가 없으며 마지막으로 누구나 배터리를 포함하는 상업적화된 다른 장치에 비해 편안하게 휴대할 수 있다.
- Author(s)
- MUHAMMAD ALI SYED
- Issued Date
- 2020
- Awarded Date
- 2020. 2
- Type
- Dissertation
- Publisher
- 부경대학교
- URI
- https://repository.pknu.ac.kr:8443/handle/2021.oak/23734
http://pknu.dcollection.net/common/orgView/200000294772
- Affiliation
- Pukyong National University, Graduate School
- Department
- 대학원 전자공학과
- Advisor
- Wan-Young Chung
- Table Of Contents
- 1. Introduction 1
1.1 Idea for the Motivation 2
1.2 Challenges in the Previous Research 3
1.3 Core Objectives of our Research 3
1.4 Chapters Organization 4
2. Background and Related Work 6
2.1 Ambient Energy Sources: Benefits and Drawbacks 6
2.2 RFID based Energy Harvesting: Benefits and Drawbacks 10
2.3 NFC based Smartphone Energy Harvesting: A New Approach 13
3. Theory on Skincare Support 17
3.1 Skincare Health Process 17
3.2 Common Skin Moisture Measurement Devices: Possible Demerits 19
3.3 Skincare Support using Skin Moisture Sensor 21
3.4 Skincare Support using UV Sensor 22
4. NFC Antenna Structure and Communication Protocol 27
4.1 Structure of the NFC Coil embedded in Smartphone 28
4.2 NFC Coil for the Skincare Device 34
4.3 Performance Simulations between Transmitting and Receiving Coils 36
4.3.1 Mutual Coupling between the NFC Smartphone Coil and Skincare Device Coil 36
4.3.2 Coupling Coefficient between the NFC Smartphone Coil and Skincare Device Coil 37
4.3.3 Magnetic Field and Energy Transfer between the NFC Smartphone Coil and Skincare Device Coil 38
4.4 Data Communication between NFC Smartphone and Skincare Device 41
5. System Model Design and Implementation 46
5.1 System Structure Overview 46
5.2 Skincare PCB Design 47
5.2.1 Circular Based PCB Design Structure 47
5.2.2 Flexible 3D Print Design 48
5.3 Android Application Design 50
5.3.1 Skin Moisture Measurement Activity 50
5.3.2 UV Irradiance Measurement Activity 51
6. Experimental Results and Discussion 52
6.1 Energy Harvesting by Skincare Device from the NFC Smartphone 52
6.2 I2C Communication between the Components of Skincare Device 53
6.3 Results of Skin Moisture: Indoor and Outdoor Environments 54
6.3.1 History of Skin Moisture in Android Application: Indoor and Outdoor Environments 57
6.4 Other Results Based on Skin Moisture Sensing Values 60
6.4.1 Transepidermal Water Loss Measurements 60
6.4.2 Skin Wetness Factor Measurements 67
6.5 UV Irradiance Results 73
7. Deep Learning for Skincare Data 76
7.1 Overview of Deep Learning Models 77
7.1.1 Feedforward Neural Network 78
7.1.2 Convolution Neural Network (CNN) 78
7.1.3 Recurrent Neural Network (RNN) 79
7.2 Output Predictions of Skincare Data 80
7.2.1 Skin Moisture Analysis using Regression Learning 80
7.2.2 Combined Six Channel Output Predictions using Artificial Neural Network(ANN) Model 82
7.3 Classification using ANN Model 85
8. Conclusion 89
- Degree
- Master
-
Appears in Collections:
- 대학원 > 전자공학과
- Authorize & License
-
- Files in This Item:
-
Items in Repository are protected by copyright, with all rights reserved, unless otherwise indicated.