Automotive Reminding System of Speed LimitBased on Vision Technique
碩士 === 大同大學 === 資訊工程學系(所) === 98 === In the researches on the automotive safety assistant system, traffic signs detection and recognition have get more and more attentions. In this paper a reminding system based on detection and recognition of speed limit traffic sign is presented. At first, a regi...
Main Authors: | , |
---|---|
Other Authors: | |
Format: | Others |
Language: | zh-TW |
Published: |
2010
|
Online Access: | http://ndltd.ncl.edu.tw/handle/11845128177269546096 |
Summary: | 碩士 === 大同大學 === 資訊工程學系(所) === 98 === In the researches on the automotive safety assistant system, traffic signs detection and recognition have get more and more attentions. In this paper a reminding system based on detection and recognition of speed limit traffic sign is presented.
At first, a region of interest is chosen then the traffic signs are extracted for their color characteristics. The contour of candidate object with noise removed is used to detect the ellipse. Unnecessary parts are removed using the geometric characteristics and the regions for identifying are segmented. These regions are normalized to a fixed size and changed into bi-level images. The bi-level images are compared with the standard sample of speed limit sign to get the limit.
According to the relative speed between the speed limit and car speed, the reminding signal are decided to remind the driver or not. To avoid the disturbance to driver the low confidence statuses are filtered. The filtered traffic signs are shown to driver to make final decision. This proposed mechanism would help to the development of intelligent Driver Support System.
|
---|