The security in road traffic has become a concern for the society, which motivating the development of Advanced Driver Assistance Systems (ADAS). An example of ADAS is the detection system for traffic signs, which aims to extract traffic information through computer vision te- chniques. In this work it was developed a system capable of detecting Brazilian traffic signs, using for this a new algorithm to filter objects. Firstly, the algorithm extracts objects based on color, applied through a threshold on the space Y’CbCr. After, the post-processing step, a neural network was used with the Fourier descriptors to isolate objects with typical shapes of the Brazilian traffic signs.
- Projects 5
- Followers 270
81100 Caserta, Province of Caserta, Italy