Hits:

Date of Publication:2012-01-01

Journal:大连理工大学学报

Affiliation of Author(s):运载工程与力学学部

Volume:52

Issue:6

Page Number:908-913

ISSN No.:1000-8608

Abstract:The road traffic signs recognition in open environment is very important
   in the field of the auxiliary safe driving, the autonomous vehicle
   navigation and so on. Based on the characteristics of road traffic
   signs, a control-point information obtaining method is adopted to
   acquire the trans-matrix of perspective projection. Then, the Hough
   transform is adopted to detect the regional model of lane line based on
   scene reconstruction. The road signs are divided with maximum
   one-dimensional information entropy division and edge features of signs
   are detected by Canny edge detector. Finally, an improved Hu moment
   invariants method is adopted to recognize the road traffic signs.
   Experimental results show that the proposed method for the road traffic
   signs recognition in open environment has good reliability and
   robustness.

Note:新增回溯数据