The invention relates to a method and a device for visualising the surroundings of a vehicle, especially for detecting and representing the traffic situation faced by a driver, by means of a night vision information system. According to said method, largely isochronous video images that are mostly from the same location but different spectral regions are amalgamated, and the different spectral regions are provided with weighting factors during the amalgamation. The video images are first weighted with a preliminary weighting factor in round figures, and are then subjected to an object analysis by means of an object identification system, and the weighting factors are locally adapted in an object-selective manner on the basis of the object analysis.