Breaking water waves still represent a very active research area, as they are involved in a broad spectrum of situations ranging from safety at sea to climate change. The detailed knowledge of the location of the free surface in case of breaking waves or wave impacts could prevent structural damages, increase the efficiency of ship hulls and improve the understanding of gas exchanges between the ocean and the atmosphere. Several state-of-the-art numerical and experimental studies rely on qualitative or intrusive, punctual evaluation of the free surface location. Agreement is found in the relevant literature regarding the need for alternative, quantitative methods in free surface measurements. With the intention of filling this gap, the goal of this thesis consists in the dense, two- and three-dimensional, experimental measurement and reconstruction of the free surface during a large-scale breaking dam flow impact with a vertical wall. The breaking wave event is recreated in a purposely designed acrylic 821 x 130 x 280 mm tank with a sliding gate system and recorded in multiple locations by a pair of identical and synchronized sensors in stereo configuration, and an additional side camera. The experimental campaign of this project consists of 71 runs in which different aspects of the breaking dam event are recorded and different initial conditions and fluid properties are tested. A repeatability analysis of the main flow structures is carried out successfully. A preliminary numerical study supports the comparison between the results obtained with the present setup and data taken from the literature. All the cameras are calibrated with standard procedures and sub-pixel accuracy in the root-mean-square directional reprojection error is achieved. Computer vision techniques are adopted for the quantitative measurements. The two-dimensional free surface profile is obtained applying a combination of motion saliency and contour detection on the frames extracted from the side camera's recordings. The dense point clouds representing the three-dimensional free surface in the impact region are reconstructed from the stereo pairs in the common metric reference system using WASS, an open-source stereo processing pipeline. Dense and quantitative data regarding the location of the free surface in space and time is successfully obtained. Agreement is found comparing the 2D and 3D experimental measurements. Furthermore, the numerical method ComFLOW is used for the numerical simulation of the same breaking dam flow and is validated with the experimental results obtained in this project. The numerical and experimental results agree overall, though differences can be observed in the shape of the free surface during the breaking of the impact wave and the consequent formation of the secondary wave. Additionally, good agreement is found comparing the present results with punctual (numerical and experimental) free surface measurements taken from the relevant literature. The results obtained are unique as similar continuous, quantitative, large-scale, dense and three-dimensional laboratory measurements of breaking waves and wave impacts have not yet been presented in the scientific literature, to the knowledge of the author at the time of writing this thesis. This work is considered to give an important contribution to the future validation of numerical methods for the study of large-scale wave impacts.