Real-Time Traffic Sign Detection and Recognition using CNN

• Autor(a) principal: Santos, Daniel Castriani
• Data de Publicação: 2020
• Outros Autores: Silva, Francisco Assis Da, Pereira, Danillo Roberto, Almeida, Leandro Luiz De, Artero, Almir Olivette [UNESP], Piteri, Marco Antonio [UNESP], Albuquerque, Victor Hugo
• Tipo de documento: Artigo
• Idioma: por
• Título da fonte: Repositório Institucional da UNESP
• Texto Completo: http://dx.doi.org/10.1109/TLA.2020.9082723; http://hdl.handle.net/11449/201734
• topic: Computer Vision; Convolutional Neural Network; Region-Based Convolutional Neural Network

Traffic signs presents on streets and highways have a distinct set of features which may be used to differentiate each one from each other. We propose in this paper a real-time traffic sign detection and recognition algorithm using neural networks. In order to detect traffic sign we used a Faster R-CNN (Region-Based Convolutional Neural Network), and to classify we used a Convolutional Neural Network using two different architectures. Some factors can make it difficult, such as light, occlusion, blurring, and others. This work can be applied in several areas, such as Advanced Driving Assistant System and autonomous cars.