查詢結果分析
來源資料
相關文獻
- 基於一般性環境偵測的自動車之電腦視覺導航
- 高速公路路況即時資訊系統之偵測車數量需求分析
- 自動車於直行種植環境之導引作業
- Vision-Based Guidance for Autonomous Land Vehicle Navigation in Outdoor Road Environments with Static and Moving Cars
- 高速公路路況即時資訊系統之偵測車數量需求分析
- Intelligent Control Theory in Guidance and Control System Design: An Overview
- 先進公車系統自動車輛定位與通訊技術之評估研究
- 一種智慧型PID控制器的實現
- 雙層隔振系統於寬頻週期負荷之主動控制
- 論自動車床凸輪製造CAM
頁籤選單縮合
題名 | 基於一般性環境偵測的自動車之電腦視覺導航=A Universal Environment Detection Approach to the Computer Visual Guidance of Autonomous Land Vehicles |
---|---|
作者姓名(中文) | 范欽雄; 簡榮富; | 書刊名 | 技術學刊 |
卷期 | 12:2 1997.06[民86.06] |
頁次 | 頁353-363 |
分類號 | 448.94 |
關鍵詞 | 電腦視覺導航; 自動車; 特徵對應; 距離估測; 智慧型控制; Computer visual guidance; Autonomous land vehicle; Feature correspondence; Distance estimation; Intelligent control; |
語文 | 中文(Chinese) |
中文摘要 | 本文提出了一個可利用周遭環境之邊緣特徵當作行進依據的電腦視覺導航系統。 在航行之前,此系統必須先輸入環地平線所在的影像水平線位置與參考目標,再利用連續影 像之特徵的位移量來計算環境距離。為了節省計算時間及減少雜訊干擾,本電腦視覺導航系 統對於每張影像只取所設定的水平線資料,並隨即用中序濾波器與低通濾波器予以處理。接 著對這種前後影像的一維資料進行適當的分段,而以啟發式的搜尋方法找出影像特徵的最合 適對應關係,再根據它們的位置、攝影機的焦距與前進距離求算出自動車的偏航角度與影像 中所對應的環境距離。最後,利用目標位置的座標值,並配合目前左、右動力輪回授的速度 ,透過可學習的控制矩陣對左、右馬達下達速度命令,以驅使自動車安全地邁向事先設定的 目標。本文所提出的方法皆在實際構建的自動車上獲得成功的驗證;在靜態的導航實驗方面 ,可以正確判斷偏航的角度,並約略計算出攝影機與環境之間的距離,而在動態的導航實驗 方面,此自動車能以智慧型的控制方式自動航向目標達 50 公尺以上。 |
英文摘要 | This paper presents a computer visual guidance system for autonomous land vehicles (ALVs) based on universal environment detection. Before guiding the ALV, both the horizon and the target of the navigation environments appearing in the captured image must first be inputted to the computer system which estimates the surrounding distances from each pair of sequential images. To save computation time and reduce noise disturbance, the computer system only accesses the range image of the horizon, which is further processed by median and lowpass filters. Also, each of the two consecutive range images of the horizon is appropriately segmented into a set of pieces for feature correspondence via a heuristic search approach. Referring to the feature locations, camera focal length, and ALV moving distances, the deviation angle and surrounding distances are then computed. Finally, a learnable control matrix, in terms of the target location and the speed difference of driving wheels, is developed for issuing the speed orders for the left and right motors to safely guide the ALV toward the target. many successful experiments have been made on an actual ALV to verify the effectiveness of the proposed methods. With respect to the static navigation experiments, the deviation angle can be accurately calculated and the surrounding distances from the camera can be approximately estimated. As regards the dynamicnavigation experiments, the ALV is guided by the intelligent control to approach a target more than 50 meters away. |
本系統之摘要資訊系依該期刊論文摘要之資訊為主。