Extraction and Recognition of Road Sign Region from Natural Color Image, by using novel HSL and Coordinates transformation

Abstract

An image processed recognition ratio of road signs as high as 97% has been successfully achieved by using novel HSL and coordinates transformation. The purpose of this research is to construct algorithm that recognizes certain road signs on roads by the image processing. A digitized color image is transformed into a color-phase image by the HSL transformation. It is then converted into a binary image, using n×n matrix filter for noise reduction. Each pixel in an extracted binary image is allocated seven boundary patterns. The road sign is, then, extracted and converted by a coordinates transformation, using the Complex-log mapping method to remove influence of expansion, reduction and rotation. The matching is, also, carried out by using the Fourier transformation to remove an influence of difference in position.