Sequential Monte-Carlo estimation based road region segmentation considering perspective projection

Road region is an important information for guidance of such objects like robots or autonomous vehicles. The goal of this research is to develop a robust monocular algorithm for extraction of road region from image sequences, captured by camera mounted on the moving objects. The key idea introduced in the first stage of this research is to express the road region in terms of probability density function (p.d.f.) of color and spatial values, and recursively estimate the p.d.f. by sequential Monte-Carlo estimation. This paper describes an extension to the proposed method in the sense that newly consider perspective projection of road plane by camera. The proposed approach is examined on real image sequences, and compared with the first one.