[n this paper, we propose a vertical and tloor line-based monoclllar simultaneous localization and mapping (SLAM) system which utilizes vertical Iines, tloor lines, and vanishing points as sensory input to perfom1 robllst SLAM in corridor environments. By combining three map featllre types, ollr design can help a robot to perfonn accurate pose estimation, repeatable loop closure, and to constrllct a more expressive environmental map. As a primitive element of a geometric structure, a line segment has one additional dimension compared to a point feature, thereby allowing the use of line segments to easily represent a geometric structllre using a smaller number of features. This system presents map features on a 2D ground space: the vertical line as a projection point, the tloor line as the original line, and the vanishing point as a directional vector. Although the vertical line, tloor line, and vanishing point use different parameterization and initialization methods, their measurement models are integrated into a unified extended Kalman filter (EKF) framework. Experimental results show that our system can be deployed in a structured indoor environment as a suitable SLAM solution.