A method for creating an object mask for video surveillance systems

The proposal is for a process for preparing an object mask for video monitoring systems by means of which spatial objects can be clearly distinguished from a background image which lies substantially in one plane. Here, the region to be monitored is imaged by two spatially separated video cameras (4, 5). A ten-parameter image of one video image (2) is undertaken in the co-ordinated system of the other corresponding other video image (1). The projected video image (8) is then compared with the corresponding other video image (11).