This paper is concerned with the enhancement of the visual feedback for the teleoperator of an all-terrain fast mobile robot. Indeed, remote control trials have shown inferior mobility performances (as speed, accuracy) in relation to direct driving of the same vehicles. Our goal of improving these performances for an operational application needs notably an better visual perception. This article deals with the first step of this goal. Studies showed notably that the teleoperator needs a large field of view. We decide to take advantage of the rough movement of the camera induced by the robot motion on unstructured ground. We use this environment sweep of the camera to dynamically build an image mosaic which is a way for extending the operator field of view using one camera. The image mosaic is made of the successive video images aligned together to display an coherent and stable environment in spite of the camera motion. The correct alignement is realized using an iterative multiresolu...