This paper presents a novel approach for real-time mobile robot localization in structured indoor environments. The proposed method takes advantage of the available structural information by implementing a Monte Carlo Localization strategy over a map of line segments rather than a grid-based map, thus allowing for speed improvements. Another novel aspect is in the likelihood function, which is based on a Modified Hausdorff Distance between the expected line segments the robot should sense and the line segments extracted from actual measurements using a range finder. Additionally, the number of particles of the Monte Carlo Localization method is automatically adjusted, using a large number of particles in the global localization phase, where the position of the robot is unknown, and a reduced number of particles during the state tracking phase, where uncertainty about the robot position is restricted to a small area. The proposed approach has been implemented and tested in a real office environment, achieving true real-time performance. Results show a fast convergence from global localization to state tracking, as well as, robustness in position tracking. Experimental tests and comparisons with state-of-the-art methods validate the efficiency and robustness of our approach