Showing vehicles at blind corners from mixed-dimensional multi-view geometry

Abstract

In this paper, we propose a method for assisting drivers by showing vehicles at blind corners. Recently, vehicles are equipped with various kinds of sensors, such as 2D cameras and 3D laser radars. In this paper, we propose a method which enables us to combine different dimensional sensors, such as 2D sensors and 3D sensors, for showing blind corner information efficiently. In particular, we consider the case where one vehicle is equipped with a 3D laser radar and another vehicle is equipped with a 2D camera, and propose a method for showing the position of invisible vehicles by combining these sensor data.