應用於自動駕駛系統中具有路標辨識功能之車道跟隨控制設計

Lane Following Control Design with Road Sign Recognition for Self-Driving Systems

王于恩(Yu-En Wang)

指導教授 ： 林容杉

國立暨南國際大學/科技學院/電機工程學系/碩士(2019年)

https://doi.org/10.6837/ncnu201900133

摘要

本文提出了一種智慧自動駕駛系統，用於道路標誌識別的自動車道跟踪。
所提出的智能自動駕駛車輛平台包含網絡攝影機和三個紅外傳感器，用於收集
有用信息。利用所獲得的信息，系統可以識別實現車道跟隨目標的道路標誌，
避障。一般來說，這種自驅動系統的結構大致可分為兩部分：傳感和決策。

在傳感部分的工作中，系統主要處理從攝影機圖像和紅外光獲得的信息。
基於檢測到的車道標記，確定車輛的相對位置以查看其是否在其車道內。在決
策方面，我們提出的自動駕駛車輛可以遵循根據模糊控制策略的期望車道。當
遇到道路標誌時，圖像處理技術具有確定道路標誌的指示的潛力，以便立即做
出適當的動作。

結果，智能車輛可以通過所提出的自驅動系統有效地操縱，以穩定地保持
在適當的車道中。此外，車輛可以成功地識別道路標誌並在遇到道路標誌時適
當地響應。給出了一些實際結果，說明我們提出的自動駕駛系統確實可以通過
道路標誌識別實現車道跟隨目的。

關鍵字

自動駕駛系統 ； 車道跟隨 ； 路牌辨識 ； 避障

並列摘要

This thesis presents an intelligent self-driving system for the purpose of automatic
lane following with road sign recognition. The proposed intelligent self-driving vehicle platform contains a web camera and three infrared sensors for the collection of useful information. With the obtained information, the self-driving system can recognize the road sign to achieve lane following objective with obstacle avoidance. Generally speaking, the structure of this self-driving system can be roughly divided into two parts: sensing and decision making.

In the works of sensing part, the system mainly deals with the information obtained from the camera images and the infrared light. Based on the detected lane markings, the relative position of vehicle is determined to see whether it is within its lane. In the part of decision-making, our proposed self-driving vehicle can follow the desired lane according to fuzzy control strategy.

When encountering a road sign, the image processing technology has the potentials
to determine the indication of the road sign in order to attain appropriate
response immediately.

並列關鍵字

Self-driving system ； Lane following ； Road sign recognition ； Obstacle avoidance

參考文獻

[1] A. Gonzlez, L. M. Bergasa and J. J. Yebes, “Text detection and recognition on

traffic panels from street-level imagery using visual appearance," IEEE Transactions on Intelligent Transportation Systems, pp. 228-238, Feb., 2014.

[2] H. Luo, Y. Yang, B. Tong, F. Wu, and B. Fan, “Traffic sign recognition using a multi-task convolutional neural network,"IEEE Transactions on Intelligent

Transportation Systems, pp. 1100-1111, Apr., 2018.

[3] N. P. Botekar and M. N. Mahalakshmi, “Development of road sign recognition