The main objective of this thesis is to develop a system for monitoring the close environment of a level crossing. It aims to develop a perception system allowing the detection and the evaluation of dangerous situations around a level crossing. To achieve this goal, the overall problem of this work has been broken down into three main stages. In the first stage, we propose a method for optimizing automatically the location of video sensors in order to cover optimally a level crossing environment. This stage addresses the problem of cameras positioning and orientation in order to view optimally monitored scenes. The second stage aims to implement a method for objects tracking within a surveillance zone. It consists first on developing robust algorithms for detecting and separating moving objects around level crossing. The second part of this stage consists in performing object tracking using a Gaussian propagation optical flow based model and Kalman filtering. On the basis of the previous steps, the last stage is concerned to present a new model to evaluate and recognize potential dangerous situations in a level crossing environment. This danger evaluation method is built using Hidden Markov Model and credibility model. Finally, synthetics and real data are used to test the effectiveness and the robustness of the proposed algorithms and the whole approach by considering various scenarios within several situations. This work is developed within the framework of PANsafer project (Towards a safer level crossing), supported by the ANR-VTT program (2008) of the French National Agency of Research. This project is also labelled by Pôles de compétitivité "i-Trans" and "Véhicule du Futur". All the work, presented in this thesis, has been conducted jointly within IRTES-SET laboratory from UTBM and LEOST laboratory from IFSTTAR.
