Title :
Identification of road signs using a new ridgelet network
Author :
Yang, Shuyuan ; Wang, Min
Author_Institution :
Dept. of Electr. Eng., Inst. of Intelligence Inf. Process., Xi´´an, China
Abstract :
Identification of road signs is a hotspot in the visual vehicle-navigation system. In this paper, a directional multiresolution ridgelet (DMRR) network is proposed based on a binary ridgelet frame to recognize traffic road signs in atrocious conditions. A detailed designing of the network and its learning algorithm are given. For the high efficiency of ridgelet in dealing with the directional information and the superiority in high dimension, this new network can obtain higher recognition rate, faster learning and has a more flexible structure than other traditional neural networks.
Keywords :
driver information systems; image recognition; learning (artificial intelligence); navigation; neural nets; DMRR network; atrocious conditions; binary ridgelet frame; directional multiresolution ridgelet network; learning algorithm; recognition rate; road sign identification; traffic road signs; visual vehicle-navigation system; Biological neural networks; Information processing; Intelligent networks; Intelligent vehicles; Multiresolution analysis; Neural networks; Neurons; Radar signal processing; Roads; Signal resolution;
Conference_Titel :
Circuits and Systems, 2005. ISCAS 2005. IEEE International Symposium on
Print_ISBN :
0-7803-8834-8
DOI :
10.1109/ISCAS.2005.1465413
