Proceeding of the Joint International Symposium on Deformatuon Monitoring. 2-4 November, 2011. Hong Kong. China. Report. Significant amounts of resources have been invested into the design, construction and maintenance of various large-scale man-made linear features around the world. However, the deformation monitoring of such constructions has become of great importance, not only for the apprehension of the financial investment loss or the efficiency of the transportation system, but mainly for the safety of human lives. Monitoring the deformation of various linear constructions (road/railways, tunnels, subways, etc) has always been a great challenge. Since aerial and satellite methods cannot provide overall solutions, especially for underground constructions, land-based systems have been also called to address the problem. Recently, Mobile Mapping Systems (MMS) have become of great interest, since they deliver kinematic surveying data of high quality while keeping the cost at low levels. This paper proposes a new method for the automatic feature extraction of the trajectory path, based on machine learning principles. It aims to deliver reliable information on the platform's navigation solution, the core of a land-based mobile system success, destined for the deformation monitoring of transportation constructions. More specifically, a point classification algorithm is proposed, in which the point fix locations of the moving vehicle are assigned to appropriate linear geometric elements. This approach relies on Least Squares Support Vector Machines (LS-SVM) for pattern recognition. Its overall scope is to provide a tool for the automatic extraction of a vehicle's trajectory and in the sequence, to deliver accurate identification of the alignment transition points. In fact, the accurate detection of the transition points is crucial, as it creates matching constrains between the construction's as-built plans and the mapping information. Consequently, direct comparisons between design cross sections and measured profiles of the construction's deck can be applied, so that critical deformation can be detected. Validation of the results and conclusions are based on real-world tests.