Proceeding of the Joint International Symposium on Deformatuon Monitoring. 2-4 November, 2011. Hong Kong. China. Report. Terrestrial laser scanners are high-accuracy 3D imaging instruments that are capable of measuring deformations with sub-millimetre level accuracy in most close-range applications. Traditionally, deformation monitoring is performed using distinct signalized targets. In this case, the centroids of these targets must be determined with great accuracy. A least-squares based target centroid extraction algorithm for planar checkerboard/chessboard targets is proposed for laser scanner data. These targets are used in a free-station network for performing deformation analysis with no assumptions about the deformation pattern. To ensure the optimum measurement accuracy of the instrument, all systematic errors inherent to the instrument at the time of data acquisition need to be removed. One method for reducing these systematic errors is self-calibration of terrestrial laser scanners. In this paper, this was performed on-site to model the systematic errors in the raw observations of the scanner. Post self-calibration, the accuracy of the measured translation movements were improved from the millimetre level to the sub-millimetre level. Despite the outstanding performance of using laser scanners with signalized targets in deformation analysis, the main benefit of active sensors like terrestrial laser scanning systems is its ability to capture 3D information of the entire scene without markers. A new markerless deformation analysis technique that utilizes intersection points derived from planar-features is proposed and tested in this paper. The extraction and intersection of planes in each point cloud can be performed automatically or semi-automatically. This new method is based on free-stationing and does not require a priori knowledge about stable control points. It can detect and measure translational and rotational movements of the planes with minimal human interaction. This paper will present both simulated and real results, demonstrating the performance of the newly proposed methodology.