Using Block-Matching Methods to Help Navigating Visually Impaired People using Raspberry PI Platform

The paper analyses the influence of various parameters in the detection and recognition of traffic signs, to help navigating people with visual impairments. We used the Raspberry PI 4 model B + platform and functions from the OpenCV library to reach our objective. 3 template matching functions from OpenCV were compared and we used the Normalized Cross Correlation (NCC or CCORR), Normalized Inter-correlation (NIC or CCOEFF) and Sum of Squared Differences (SSD or SQDIFF) as matching metrics from the class of intensity-based methods for template matching. The goal was to identify what method gives the best results for the images with traffic signs taken in the evening.