JICCE (한국정보통신학회)
Current Result Document :
| 한글제목(Korean Title) |
A Method for Improving Accuracy of Image Matching Algorithm for Car Navigation System |
| 영문제목(English Title) |
A Method for Improving Accuracy of Image Matching Algorithm for Car Navigation System |
| 저자(Author) |
Jin-Deog Kim
Hye-Young Moon
|
| 원문수록처(Citation) |
VOL 09 NO. 04 PP. 0447 ~ 0451 (2011. 08) |
한글내용
(Korean Abstract) |
|
영문내용
(English Abstract) |
Recently, various in-vehicle networks have been developed respectively in order to accomplish their own purposes such as CAN and MOST. Especially, the MOST network is usually adapted to provide entertainment service. The car navigation system is also widely used for guiding driving paths to driver. The position for the navigation system is usually acquired by GPS technology. However, the GPS technique has two serious problems. The first is unavailability in urban canyons. The second is inherent positional error rate. The problems have been studied in many literatures. However, the second still leads to incorrect locational information in some area, especially parallel roads.
This paper proposes a performance tuning method of image matching algorithm for the car navigation system. The method utilizes images obtained from in-vehicle MOST network and a real-time image matching algorithm which determines the direction of moving vehicle in parallel section of road. In order to accuracy improvement of image matching algorithm, three conditions are applied. The experimental tests show that the proposed system increases the accuracy. |
| 키워드(Keyword) |
Car Navigation System
Image Matching
Performance Tuning
GPS Alternatives
|
| 파일첨부 |
PDF 다운로드
|
