Hardware architecture: Correlation-based approach for road sign detection
Sana Bougharriou, Fayçal Hamdaoui, Abdellatif Mtibaa · 2016
Road Sign Detection (RSD) is becoming a major goal of the safety Advanced Driving Assistance Systems (ADAS). Automotive research area share many publications based various techniques used to detect and classify signs. This paper provides a hardware detection-based correlation architecture using Xilinx System Generator (XSG). This proposed architecture outsets with pre-processing step: RGB to YCrCb space, thresholding and closing (erosion and then dilation). Then signs are classified using an intelligent technique: the correlation method. Experimental results are demonstrated using the proposed architecture applied on a set of traffic signs stored in a database, and then compared to software results to conclude about the the quality and the efficiencies of this architecture.