EFFICIENT LANE DETECTION BASED ON ARTIFICIAL NEURAL NETWORKS

• DOI: 10.5194/isprs-annals-iv-4-w3-13-2017
• Publication Date: 2017-09-26T10:13:35Z
• Authors: F. Arce; E. Zamora; G. Hernández; H. Sossa
• Keywords: Technology; T; 11. Sustainability; 0202 electrical engineering, electronic engineering, information engineering; Applied optics. Photonics; 02 engineering and technology; TA1-2040; Engineering (General). Civil engineering (General); TA1501-1820

Abstract. Lane detection is a problem that has attracted in the last years attention of computer vision community. Most approaches used until now to face this combine conventional image processing, analysis and pattern classification techniques. In paper, we propose methodology based on so-called Ellipsoidal Neural Networks with Dendritic Processing (ENNDPs) as new approach provide solution important problem. The functioning performance proposed validated real video taken by camera mounted car circulating urban highway Mexico City.