| 研究生: |
歐軒慈 Hsuan-Tzu Ou |
|---|---|
| 論文名稱: |
基於慣性感測器與肌肉訊號之穿戴式裝置三維手寫身份認証 An Wearable with IMU and EMG sensors for 3D space handwritting authentication |
| 指導教授: |
范國清
Kuo-Chin Fan |
| 口試委員: | |
| 學位類別: |
碩士 Master |
| 系所名稱: |
資訊電機學院 - 資訊工程學系 Department of Computer Science & Information Engineering |
| 論文出版年: | 2016 |
| 畢業學年度: | 104 |
| 語文別: | 中文 |
| 論文頁數: | 78 |
| 中文關鍵詞: | 慣性感應器 、肌肉訊號 、手環穿戴裝置 、身份認証 |
| 外文關鍵詞: | IMU, EMG, MYO Armband, Authentication |
| 相關次數: | 點閱:10 下載:0 |
| 分享至: |
| 查詢本校圖書館目錄 查詢臺灣博碩士論文知識加值系統 勘誤回報 |
在數位化充斥著生活的現在,人們的生活幾乎已離不開電腦手機等資訊產品,在此時空下傳統的密碼金鑰等資訊認證或保密的方法,已經難以應付目前的需求。目前基於生物行為特徵,所發展出的認證方法已廣為大眾接受,此種認證方式是彌補傳統密碼金鑰易遭複製易被遺忘等缺陷的方法。
目前可當作身份認證的生物特徵有很多,本論文以三維空中手寫簽名與肌肉訊號當作個人辨識特徵,資料獲取方式為自製慣性感應器穿戴裝置,裝置於手指部位獲取手指姿態,並使用 Thalmic labs 的 MYO Armband 取得手臂肌肉訊號與手臂姿態,利用三點定位的方式,計算三維空中手寫軌跡作為部分手寫特徵,最後使用循環神經網路訓練身份認證系統。
We have lived in a digital life nowadays. We cannot live without computers, smart phone and other information products in our daily life. Those traditional confidentiality and authentication methods, for example cryptographic keys, is too old to cope with the new problems in recent years. Currently, some authentication methods based on biological behavioral characteristics have been widely accepted. Such certification procedures find a new way to help the traditional cryptographic keys which would be copied and forgotten easily.
There are many methods for biometric characteristics. The proposed method uses handwritten signals in a 3D coordinate space and EMG signals as personal identification features. We use an IMU wearable device which is made by ourselves and equiped on fingers to get the pose signals of fingers. Then, we use MYO Armband from Thalmic labs to get EMG and Arm altitude signals and use three-points fix method to calculate handwritten trajetory as part of the features. Finally, we use recurrent neural network with LSTM to train the authentication system.
[1] Sepp Hochreiter, Jurgen Schmidhuber. Long Short-Term Memory. Neural Computation 9(8):1735-1780, 1997.
[2] Alex Graves. Supervised Sequence Labelling with Recurrent Neural Network. http://www.cs.toronto.edu/~graves/
[3] P. J. Werbos. Backpropagation through time: what it does and how to do it. Proceedings of the IEEE, VOL, 78 NO. 10, Oct 1990
[4] Wojciech Zaremba. Recurrent Neural Network Regularization. ICLR 2015.
[5] Sebastian O.H. Madgwick, Andrew J.L. Harrison, Ravi Vaidyanathan. Estimation of IMU and MARG orientation using a gradient decent algorithm. 2011 IEEE International Conference on Rehabilitation Robotics. http://www.x-io.co.uk/open-source-imu-and-ahrs-algorithms/
[6] H. Sakoe ; Nippon Electric Company, Limited, Kawasaki, Japan ; S. Chiba. Dynamic programming algorithm optimization for spoken word recognition. IEEE Transactions on Acoustics, Speech, and Signal Processing VOL, 26 NO, 1, pp43-49, Feb 1978.
[7] Tu-Liang Hsu, Cheng-Ling Chu, Yi-Ju Tsai, and Jeen-shing Wang, Member, IEEE. An Inertial Pen With Dynamic Time Warping Recognizer for Handwriting and Gesture Recognition. IEEE SENSORS JOURNAL, VOL, 15 NO. 1, JANUARY 2015.
[8] 蘇木春, 張孝德. 機器學習 類神經網路、模糊系統以及基因演算法 修訂二版. 全華科技圖書股份有限公司, 2012, ISBN 957-21-4737-4.
[9] InvenSense. MPU-6000 and MPU-6050 Register Map and Descriptions Revsion 4.0.
[10] Werbos, Bishop. multilayer perceptron.
[11] Jeff Elman. Finding Structure in Time. CONGNITIVE SCIENCE 14, 179-211 1990.
[12] Stan Salvador, Philip Chan. FastDTW: Toward Accurate Dynamic Time Warping in Linear Time and Space. Dept. of Computer Sciences Florida Institute of Technology Melbourne, FL 32901.