• 沒有找到結果。

利用虛擬實境模擬系統偵測駕駛員從清醒至打瞌睡過程之腦波變化

N/A
N/A
Protected

Academic year: 2021

Share "利用虛擬實境模擬系統偵測駕駛員從清醒至打瞌睡過程之腦波變化"

Copied!
9
0
0

加載中.... (立即查看全文)

全文

(1)

多媒體工程研究所

利用虛擬實境模擬系統偵測駕駛員從清醒

至打瞌睡過程之腦波變化

EEG-Based Subject- and Session-Independent Drowsiness

Detection: An Unsupervised Approach

研 究 生:莊玠瑤

指導教授:林進燈 教授

(2)

利用虛擬實境模擬系統偵測駕駛員從清醒至打瞌睡過程之腦波變化

EEG-Based Subject- and Session-Independent Drowsiness Detection:

An Unsupervised Approach

研 究 生:莊玠瑤 Student:Chien-Yao Zhuang

指導教授:林進燈 Advisor:Chin-Teng Lin

國 立 交 通 大 學

多 媒 體 工 程 研 究 所

碩 士 論 文

A Thesis

Submitted to Institute of MultimediaEngineering College of Computer Science

National Chiao Tung University in partial Fulfillment of the Requirements

for the Degree of Master

in

Computer Science July 2008

Hsinchu, Taiwan, Republic of China

(3)

利用虛擬實境模擬系統偵測駕駛員從清醒至打瞌睡過程之腦波變化

研 究 生:莊玠瑤 指導教授:林進燈

國立交通大學資訊工程學系多媒體研究所

中文摘要

打瞌睡是造成意外事故的主因之一,因此於各種工作環境中,一套可靠、即時的 非侵入式打瞌睡警示系統的建立是有其必要性的。本論文的目標在於利用360 度 虛擬實境(Virtual-Reality: VR)模擬駕駛系統,藉由一小時將維持車輛在車道 中 心 位 置 的 長 時 駕 駛 工 作 , 偵 測 駕 駛 員 由 清 醒 到 打 瞌 睡 的 連 續 腦 波 (Electroencephalogram: EEG)變化現象。十三位年齡在18到28歲間的受測者參 與此駕駛模擬實驗,並以250Hz 取樣頻率同步量測其28通道腦電波與駕駛行為資 料。所量測的腦電波利用 unsupervised 演算法來偵測人類從清醒到打瞌睡認知 狀態的改變。此應用可作為未來發展即時瞌睡警示系統的基礎。實驗結果顯示, 我們不需要事先資料的回饋並且使用更簡潔的運算即可準確的偵測出受測者從 清醒到打瞌睡的腦波狀態。並發現人類在不同打瞌睡的程度之下其腦電波的變化 情形也不相同。精神狀態從清醒至極輕度和輕度瞌睡過程中,有些的受測者可能 使用α波來做特徵值抽取會有比較好的表現結果,而有些受測者則可能使用θ波 會有比較好的表現結果 所以我們結合α波和θ波來做為一個瞌睡指標的依據, 以期能偵測出受測者從清醒到打瞌睡的變化來防止一些因打瞌睡而導致的意外 產生。 關鍵字︰打瞌睡、虛擬實境、腦電波、認知狀態、瞌睡警示,特徵值抽取

(4)

EEG-Based Subject- and Session-Independent

Drowsiness Detection:

An Unsupervised Approach

Student: Chien-Yao Zhuang Advisor: Dr. Chin-Teng Lin Department of Computer Science

National Chiao Tung University

Abstract

Monitoring and prediction of changes in the human cognitive stages, such as alertness, drowsiness, using physiological signals such as Electroencephalogram (EEG) are very important for driver’s safety. Typically, psychophysiological studies on real time detection of drowsiness based on EEG data use the same model for all subjects. However, the relatively large individual variability in EEG dynamics relating to loss of alertness implies that for many subjects, group statistics may not be useful to accurately predict changes in cognitive states. Researchers have attempted to build subject-dependent models based on his/her pilot data to account for individual variability. Such approaches cannot account for the cross-session variability in EEG dynamics, which may cause problems due to various reasons including electrode displacements, environmental noises, and skin-electrode impedance. Here we propose an unsupervised subject and session independent approach for detection departure

(5)

from alertness. We first demonstrate that the EEG power in the alpha band (as well as in the theta band) is correlated with changes in the subject’s cognitive state with respect to drowsiness as reflected through his driving performance. We then make a few mild and realistic assumptions to derive models for the alert state of the driver using the EEG power in the alpha and theta bands. The deviations of the EEG power in the alpha and theta bands from the corresponding alert models are found to be correlated with the changes in the driving performance. Although, the alert state models derived using alpha band power and theta band power are quite effective in detecting drowsiness, for an improved performance, we also use a liner combination of deviations of the EEG power in the alpha band and theta band from the respective alert models. This approach being an unsupervised and session independent one could be used to develop a useful system for noninvasive monitoring of the cognitive state of human operators in attention-critical settings.

Keyword: human cognition, Electroencephalogram (EEG), Alertness, Drowsiness, unsupervised, alpha band, theta band.

(6)

誌 謝

t 本論文的完成,首先要感謝我的指導教授 林進燈博士在過去兩年研究期 間,提供豐富的研究資源和實驗環境,使得本文得以順利完成。 其次,我要感謝我的父母對我的照顧與栽培,教導我做人品德為最,強調人 格健全之發展與學習生活之態度,由於他們辛勞的付出和細心的照顧,才有今天 的我。

特別感謝 Professor Nikhil R. Pal 對我研究的指導還有幫忙論文的修改,嚴謹 的研究態度,寬宏的胸襟,表裡如一,他是一位真正的學者,我永遠記得,當我 送他一座代表財富的中國風財神娃娃時,他說他最想要的禮物是 Intelligence。 感謝美國加州聖地牙哥大學的 鐘子平教授、 段正仁教授及 黃瑞松學長, 給予我研究上最大的協助,給我最專業的意見跟看法。 另外,我要感謝腦科學研究實驗室的全體成員,沒有他們也就沒有我個人的 成就。另外要感曲在雯博士、趙志峰學長、黃騰毅學長、陳玉潔學姊及柏銓、奎 銘、德正、孟修、青甫以及尚文同學在過去兩年研究生活中同甘共苦,相互扶持。 此外,我也要感謝柯立偉學長在研究上的幫助,也感謝實驗室助理在許多事務上 的幫忙。 最後,要特別感謝的人,就是這八年來辛苦陪伴我的老婆 Kristar,我的人 生因妳而有所不同,我低潮時你總是會設法使我振作,我懈怠時,妳也會適時的 鞭策我,激勵我,感謝妳的付出,而我也會以我往後的生命以及妳最最大的幸福 最為回報。 謹以本文獻給我親愛的家人與親友們,以及關心我的師長,願你們共享這份 榮耀與喜悅。

(7)

Contents

1. Introduction...1

1.1. Current researches of drowsiness...1

1.2. Kinesthetic perception during driving ...3

1.3. Virtual reality dynamic simulator ...5

1.4. Organization of this thesis ...6

2. System architecture and experimental design...8

2.1. 3D virtual reality driving simulation environment ...8

2.2. The EEG data collection ...11

2.3. Subjects ...12

3. The unsupervised approach ...14

3.1. Indirect measurement of alertness ...16

3.2. Smoothing of the power spectra ...18

3.3. Computation of the alert model of the subject...20

3.4. Computation of the deviation from the subject...24

3.5. Sorted driving performance spectral analysis ...26

4. Results ...27

4.1 The choice of channels...27

4.2 Performance sorted analysis ...29

4.3 Linear combination of model deviations ...31

5. Discussion...38

5.1 Alertness model ...38

5.2 Feature selection ...39

5.3 Supervised and unsupervised...39

5.4 Threshold from alertness to drowsiness...41

5.5 Threshold from alertness to drowsiness...42

5. Conclusion ...45

(8)

Figures

Figure 1: The dynamic VR driving environment...9

Figure 2: The task of the driving experiment ...10

Figure 3: The EEG cap and electrodes placement of international 10–20 system ...12

Figure 4: The flowchart of the unsupervised EEG analysis method ...16

Figure 5: The process of driving performance...18

Figure 6: The EEG power spectral analysis procedure...20

Figure 7: Example of MDA, MDT, and driving performance ...28

Figure 8: Performance-sorted EEG spectra at Oz over 13 sessions...30

Figure 9: Performance-sorted MDA/MDT at Oz over 13 sessions. ...31

Figure 10: MDT and MDA vs. actual driving performance.. ...34

Figure 11: MDC from Oz channel and the driving performance of all subjects ...36

Figure 12: The flowchart of the supervised EEG analysis method. ...40

Figure 13: Depend on driving performance to set threshold ...42

Figure 14: The sorted performance of MDC ...43

(9)

Tables

Table 1: The average correlation of all subjects and all channels...29

Table 2: The comparison of the correlation between different feature from OZ ...32

參考文獻

相關文件

Students are provided with opportunities to learn and develop the skills, strategies and confidence needed to participate in Guided and Independent Reading as well as the

Robinson Crusoe is an Englishman from the 1) t_______ of York in the seventeenth century, the youngest son of a merchant of German origin. This trip is financially successful,

fostering independent application of reading strategies Strategy 7: Provide opportunities for students to track, reflect on, and share their learning progress (destination). •

Strategy 3: Offer descriptive feedback during the learning process (enabling strategy). Where the

Now, nearly all of the current flows through wire S since it has a much lower resistance than the light bulb. The light bulb does not glow because the current flowing through it

To this end, we introduce a new discrepancy measure for assessing the dimensionality assumptions applicable to multidimensional (as well as unidimensional) models in the context of

• We have found a plausible model in AdS/CFT that captures some essential features of the quantum hall eect as well as a possible experimental prediction. • Even if this is the

There are existing learning resources that cater for different learning abilities, styles and interests. Teachers can easily create differentiated learning resources/tasks for CLD and