道路視覺偵測與自動導航系統之實現 ; Lane Visual Detection and Realization of the Autonomous Vision-Guided System

NCUIR > College of Electrical Engineering & Computer Science > Graduate Institute of Electrical Engineering > Electronic Thesis & Dissertation > Item 987654321/44616

Please use this identifier to cite or link to this item: http://ir.lib.ncu.edu.tw/handle/987654321/44616

Title:	道路視覺偵測與自動導航系統之實現;Lane Visual Detection and Realization of the Autonomous Vision-Guided System
Authors:	陳宏偉;Hong-Wei Chen
Contributors:	電機工程研究所
Keywords:	影像處理;模糊控制;車道線偵測;Fuzzy Control;Lane Detection;Image Processing
Date:	2010-06-29
Issue Date:	2010-12-09 13:50:36 (UTC+8)
Publisher:	國立中央大學
Abstract:	本篇論文主要為實現影像處理與控制器設計應用於輪型機器人之導航。整體架構以筆記型電腦為核心，搭配微控制器(BS2PX-IC)，整合視覺影像與馬達控制，將控制策略建構在微控制器上，實現以模糊控制為基礎的輪型機器人導航。影像處理技術利用安裝在輪型機器人前方的CCD動態拍攝前方路況，透過擷取的影像做即時影像處理後取得路面資訊，進而找出車道線位置並估測出虛擬的車道中心線作為導航的依據。本論文主要的重點分為兩部份：第一部份是影像處理及車道線偵測，第二部分是應用第一部分的結果，進行輪型機器人的導航。在影像處理及車道線偵測部份，為了因應不同的路面條件，使用動態二值化法將影像做二值化處理，取得基本車道線資訊，而根據車道線的性質將車道分為近端車道及遠端車道，在近端車道中用直線方程式去逼近以求得車道消失線(Vanishing line)，除了簡化整張影像所要處理的部份外，亦作為動態搜索車道線交點的偵測線上限值。經過多層影像處理求得車道線後再求得車道中心線以獲取導航所需之角度。第二部份是模糊控制法則，在取得道路資訊後，利用影像處理後所獲取的車道線資訊建立模糊規則庫來進行輪型機器人的導航。The goal of this thesis is to realize image processing and controller design applies to navigate of the two wheels mobile robot. The control system is implemented in a laptop, along with microcontroller (BS2PX-IC). Integrating with vision image and motor control, the control strategy is established in microcontroller to realize navigation two wheels mobile robot based on fuzzy control. Utilize the CCD camera that mounted on the front of two wheels mobile robot to take dynamic road condition. With retrieve the image to obtain efficient lane line condition and find out lane line, and then estimate the lane centre line for navigation. The concept of this thesis can be divided into two parts, the first part is image processing. The second part is using the application of the result in the first part to navigate with wheeled mobile robot. In order to adapt the different surface of lane, we use dynamic threshold to distinct black and white in lane detection. Then, according to the difference of lane, we classified into near-point-lane and far-point-lane. In near-point-lane, we use function to calculate out the vanishing line. This method simply the image processing also the line becomes the upper bound of dynamic lane detection. After multi-image processing, we solve out the mid-line of the lane and obtain the angle for navigation. The second part is fuzzy control, after finishing the image processing part, we construct fuzzy rule to navigate the wheeled mobile robot.
Appears in Collections:	[Graduate Institute of Electrical Engineering] Electronic Thesis & Dissertation

Files in This Item:

File	Description	Size	Format
index.html		0Kb	HTML	799	View/Open

社群 sharing

Loading...