Traffic Signs Recognition System with Convolution Neural Networks

摘要

The purpose of this paper is to develop an automatic traffic sign recognition system, making use of computation vision techniques and convolution neural networks. The work is divided in two phases, namely detection and classification, and here is presented a different approach on the detection phase. The tests were performed in a simulator and in a real controlled environment using the framework ROS (Robot Operating System) and implemented with the AmigoBot robot.