The use of Computer Vision techniques for the automaticrecognition of road signs is fundamental for the development of intelli-gent vehicles and advanced driver assistance systems. In this paper, wedescribe a procedure based on color segmentation, Histogram of Ori-ented Gradients (HOG), and Convolutional Neural Networks (CNN) fordetecting and classifying road signs. Detection is speeded up by a pre-processing step to reduce the search space, while classication is carriedout by using a Deep Learning technique. A quantitative evaluation of theproposed approach has been conducted on the well-known German Traf-c Sign data set and on the novel Data set of Italian Trac Signs (DITS),which is publicly available and contains challenging sequences capturedin adverse weather conditions and in an urban scenario at night-time.Experimental results demonstrate the eectiveness of the proposed ap-proach in terms of both classication accuracy and computational speed.
Fast traffic sign recognition using color segmentation and deep convolutional networks
BLOISI, Domenico Daniele
2016-01-01
Abstract
The use of Computer Vision techniques for the automaticrecognition of road signs is fundamental for the development of intelli-gent vehicles and advanced driver assistance systems. In this paper, wedescribe a procedure based on color segmentation, Histogram of Ori-ented Gradients (HOG), and Convolutional Neural Networks (CNN) fordetecting and classifying road signs. Detection is speeded up by a pre-processing step to reduce the search space, while classication is carriedout by using a Deep Learning technique. A quantitative evaluation of theproposed approach has been conducted on the well-known German Traf-c Sign data set and on the novel Data set of Italian Trac Signs (DITS),which is publicly available and contains challenging sequences capturedin adverse weather conditions and in an urban scenario at night-time.Experimental results demonstrate the eectiveness of the proposed ap-proach in terms of both classication accuracy and computational speed.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.