跳到主要內容

簡易檢索 / 詳目顯示

研究生: 李芷融
Chih-Jung Lee
論文名稱: 評估以聲音偵測為基礎的行人專用來車預警系統設計
Evaluation for designing audio-based early warning system of vehicle approaching event for improving pedestrian's safety
指導教授: 張寶基
口試委員:
學位類別: 碩士
Master
系所名稱: 資訊電機學院 - 通訊工程學系在職專班
Executive Master of Communication Engineering
論文出版年: 2017
畢業學年度: 105
語文別: 中文
論文頁數: 58
中文關鍵詞: 車輛接近聲音偵測機器學習
相關次數: 點閱:9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
  • 近年運動風氣興起,許多運動愛好者在郊區跑步或健走時亦習慣配戴耳機聆聽智慧型行動電話上播放的音樂。因聽覺封閉或分心而降低對週遭環境的感知,可能增加了與路上車輛發生事故的機會。
    為了改進行人在此情境下的安全,本論文提出了一個簡單的系統設計並評估,以便將來在智慧型行動裝置平台上實現偵測後方來車的預警系統。
    環境中的聲音將可由與行動裝置連接的指向性麥克風來收音,並經過數位訊號處理的方式取出短時間長度的音框內基本的聲音特徵:RMS(Root Mean Square)、ZCS(Zero Crossings)、SPC(Spectral Centroid)、SPR(Spectral Rolloff)。並將所取的得聲音特徵輸入至不同的機器學習分類器:kNN(K Nearest Neighbor)、MLP(Multi Layer Perceptron)、Decision Tree、Random For-est做分類,判斷該音框內是否存在來車的事件。
    本研究呈現系統的辨識率及可行性,也檢視了不適用的情境。


    More and more people tend to carry smart phones and wear headphones or headsets, listening to the music while they are jogging or walking in the suburb area. This behavior could bring distraction or temporarily losing the hearing of environmental background and cause accident to happen.
    This work proposes a simple design for audio-based early warning system of vehicle approaching event for improving pedestrian’s safety and gives evaluation.
    Sound signals were collected by an external directional microphone connected to the smart phone. Multiple feature techniques like root mean square, zero crossings, spectral centroid, and spectral rolloff were applied on the short-time frames of audio samples. Multiple machine learning classifiers like K Nearest Neighbor, Multi-layer Perceptron, Decision Tree and Random Forest were applied to classify the audio frames to detect vehicle approaching sound.
    The results showed the accuracy and the feasible of the system, also point out the circumstance can’t be applicable.

    摘要 i Abstract ii 誌謝 iii 目錄 iv 圖目錄 vii 表目錄 x 第一章 緒論 1 1-1 研究背景 1 1-2 研究目的與動機 2 1-3 論文架構 4 第二章 文獻回顧及研究方法 5 2-1 文獻回顧 5 2-2 研究方法 5 第三章 聲音特徵截取 12 3-1 Root Mean Square 12 3-2 Zero Crossings 12 3-3 Spectral Centroid 12 3-4 Spectrol Rolloff 13 3-5 都卜勒效應 15 第四章 機器學習 16 4-1 K-近鄰演算法(K Nearest Neighbors algorithm)17 4-2 多層感知器(Multilayer Perceptron) 18 4-3 決策樹(Decision Tree) 19 4-4 隨機森林(Random Forest) 20 第五章 實驗與分析討論 21 5-1 特徵統計分佈觀察 21 5-2 特徵趨勢觀察 26 5-3 實驗流程 28 5-3-1 實驗一 29 5-3-2 實驗二 30 5-3-3 實驗三 31 5-3-4 實驗四 33 5-3-5 實驗五 36 5-4 實驗結果討論 37 5-4-1 特徵縮放的影響 38 5-4-2 記億體限制 38 5-4-3 不平衡類分佈的影響 38 第六章 結論與未來展望 39 參考文獻 40 附錄一 kNN之k值於實驗三、四、五組合結果 43 附錄二 實驗二 Decision Tree輸出圖型 44

    [1] Lichenstein, Richard, et al. "Headphone use and pedestrian injury and death in the United States: 2004–2011." Injury prevention (2012): injuryprev-2011.
    [2] “NHTSA sets 'Quiet Car' safety standard to protect pedestrians”, https://www.nhtsa.gov/press-releases/nhtsa-sets-quiet-car-safety-standard-protect-pedestrians [accessed on:18 July 2017]
    [3] VK Ananthanarayanan , “Audio based detection of rear approaching vehi-cles on a bicycle”, New Brunswick, New Jersey, 2012
    [4] M.P. Paulraj, and et al. “Moving Vehicle Recognition and Classification Based on Time Domain Approach”, Procedia Engineering, Volume 53, 2013
    [5] Pillos, Angelos, et al. “A Real-time Environmental Sound Recognition Sys-tem for the Android Os.”, DCASE, 2016
    [6] “AT9913iS Specifications”, https://www.audio-technica.com.hk/in-dex.php?op=productdetails&pid=673&lang=eng [accessed on: 6 July 2017]
    [7] Nilsson, Mats E. "A-weighted sound pressure level as an indicator of short-term loudness or annoyance of road-traffic sound." Journal of Sound and Vi-bration 302.1 (2007): 197-207.
    [8] “Noise Basics and Matrics”, http://www.noisequest.psu.edu/noisebasics-ba-sics.html [accessed on: 6 July 2017]
    [9] McKay, Cory, Ichiro Fujinaga, and Philippe Depalle. "jAudio: A feature ex-traction library." Proceedings of the International Conference on Music In-formation Retrieval. 2005.
    [10] Hall, Mark, et al. "The WEKA data mining software: an update." ACM SIGKDD explorations newsletter 11.1 (2009): 10-18.
    41
    [11] Tzanetakis, George, and Perry Cook. "Multifeature audio segmentation for browsing and annotation." Applications of Signal Processing to Audio and Acoustics, 1999 IEEE Workshop on. IEEE, 1999.
    [12] Petrescu, Florian Ion. A New Doppler Effect. BoD–Books on Demand, 2012.
    [13] Dudani, Sahibsingh A. "The distance-weighted k-nearest-neighbor rule." IEEE Transactions on Systems, Man, and Cybernetics 4 (1976): 325-327.
    [14] Rumelhart, David E., Geoffrey E. Hinton, and Ronald J. Williams. "Learn-ing representations by back-propagating errors." Cognitive modeling 5.3 (1988): 1.
    [15] Quinlan, J. Ross. "Induction of decision trees." Machine learning 1.1 (1986): 81-106.
    [16] Chen, Chao, Andy Liaw, and Leo Breiman. "Using random forest to learn imbalanced data." University of California, Berkeley 110 (2004).
    [17] Friedman, Jerome, Trevor Hastie, and Robert Tibshirani. The elements of statistical learning. Vol. 1. New York: Springer series in statistics, 2001.
    [18] “BMW DRIVEBY”, https://www.freesoundeffects.com/free-track/bmw-driveby-466312/ [accessed on: 6 July 2017]
    [19] “Car driveby2”, https://www.freesoundeffects.com/free-track/car-driveby2-466317/ [accessed on: 6 July 2017]
    [20] “Car speed 01”, https://www.freesoundeffects.com/free-track/car-speed-01-466323/ [accessed on: 6 July 2017]
    [21] “Car Driveby 3”, https://www.freesoundeffects.com/free-track/carby2-466324/ [accessed on: 6 July 2017]
    [22] “Wind 2”, https://www.freesoundeffects.com/free-track/wind01-428702/ [accessed on: 6 July 2017]
    42
    [23] “Bird”, https://freesound.org/people/jmiddlesworth/sounds/364663/ [ac-cessed on: 6 July 2017]

    QR CODE
    :::