This paper is about environment perception for navigation system in outdoor applications. Unlike other approaches that try to detect an obstacle of binary state, we consider here a Digital Elevation Map (DEM). This map has to be built in regards to the guidance system's needs. These needs depend on the vehicle capabilities, its dynamics constraints, its speed etc... Starting with the navigation system's needs, our goal is to estimate a precise and dense DEM. Our approach is based on a SLAM algorithm combining a one map rangefinder and a camera. Thus we can estimate both the displacement between two laser scan and have a good density of the reconstructed DEM. The approach has been validated both using simulated realistic data and in real conditions with our robot in outdoor environement. The estimation of the DEM has been done successfully in realtime (approx. 40ms per loop).