A single video camera is installed on a vehicle body. Two still pictures of an obstacle are obtained at different vehicle traveling positions. A picture memory stores the obtained still pictures. A feature point extracting section extracts at least one feature point from the data of the stored still pictures. A three-dimensional coordinate data calculating section calculates the distance from the vehicle body to the feature point and the height of the feature point based on the positional data of the feature point on the screen and the vehicle traveling distance data. A stereoscopic picture forming section forms a stereoscopic picture image of a detected obstacle based on three-dimensional coordinate data of the feature points. A monitor displays a plan view or a side view showing the relationship between the vehicle body and the obstacle, thereby allowing a driver to directly grasp the distance of the obstacle. When the obstacle is located within a predetermined distance from the vehicle body, the driver is cautioned against a predictable collision by the alarm sound or related image.