利用聲音訊號實現車輛偵測及定位

Vehicle Tracking with Acoustic Signal

楊樂多(Le-Duo Yang)

指導教授 ： 蔡欣穆

國立臺灣大學/電機資訊學院/資訊網路與多媒體研究所/碩士(2020年)

https://doi.org/10.6342/NTU202000916

摘要

隨著現代社會中科技在各個領域中的蓬勃發展，人們期望透過設計 一套套系統來提升人類的生活品質。在日常生活的交通情境中，我們 常以各種訊號處理演算法來提升其安全性與便捷性。在這篇論文當中， 筆者將聚焦於十字路口的車輛追蹤。這些路口的交通往往會因為車輛 常常在此加速、減速、以及轉向而更為複雜。此演算法採用都普勒效 應來擷取目標的行進方向。而在距離部分，此演算法運用訊號的強度 響應結果來判斷物體的位置。在提出的系統當中，筆者以粒子濾波器 來綜合以上兩種特徵值而得到物體的即時位置座標。此外，相較於其 他使用麥克風陣列的做法，我們提出的系統在維持穩定性的同時，僅運用一個包含兩支麥克風的系統，因而降低系統的成本以及運算的複 雜度。實驗結果顯示此系統可以在誤差 1 公尺的準確度獲得目標車輛 的座標位置，而確保了此系統在實境的應用性。

關鍵字

車輛追蹤 ； 聲學信號處理 ； 粒子濾波器

並列摘要

In modern society, as technology flourishes in countless research fields, it is desirable to design systems that enhance human living qualities. In daily traffic scenarios, the safety and convenience are two main goals to achieve by the power of signal processing algorithms. In this thesis, the author focuses on vehicle tracking that takes place in a complicated crossroad scenario, where vehicle turns, brake, and speed up can constantly occur. Algorithm that con- siders Doppler effects are applied to track for the directions of the moving vehicles. Time-Frequency signal power responses are applied to obtain the distances of the moving vehicle. These two methods are then combined in the particle filter of the proposed joint position prediction system. Furthermore, unlike other works that were implemented with microphone arrays, we pro- pose a two microphone system which drastically reduces the system cost and computation effort while maintaining the system robustness. Experimental results show that our system can predict the target’s position with tracking error about one meters, which is applicable for real-world applications.

並列關鍵字

Vehicle tracking ； Acoustic signal processing ； Particle filter

參考文獻

[1]  Cross correlation. https://en.wikipedia.org/wiki/Cross-correlation.

[2]  Inverse-square law. https://en.wikipedia.org/wiki/Inverse-square_law.

[3]  S. Barnwal, R. Barnwal, R. Hegde, R. Singh, and B. Raj. Doppler based speed estimation of vehicles using passive sensor. In 2013 IEEE International Conference on Multimedia and Expo Workshops (ICMEW), pages 1–4, July 2013.

[4]  M. S. Brandstein and H. F. Silverman. A practical methodology for speech source localization with microphone arrays. Computer Speech and Language, 11(2):91 – 126, 1997.

[5]  V. Cevher, A. C. Sankaranarayanan, J. H. McClellan, and R. Chellappa. Target tracking using a joint acoustic video system. IEEE Transactions on Multimedia, 9(4):715–727, June 2007.