Method of intruder detection

Abstract

Abstract Using only a video camera (VK) and a video image processing unit (VVE), which, by a differential imaging process applied to two temporally offset video images, recognises objects (O), a predetermined surveillance zone is subjected to surveillance. The video camera (VK) exhibits a lens (OV) with a regulating device (RV) controlled by a computer ( mu C). Using the regulating device (RV), the lens parameters: focal length (f), depth of focus (S), aperture (B) and supplementary filter (F) are controlled or determined. The sighting direction of the video camera (VK) is picked up by the camera parameters: azimuth angle, polar angle and camera coordinates, by the microcomputer ( mu C). As soon as the video image processing unit (VVE) detects a moving object (O), the object (O) is sharply focused by means of the microcomputer ( mu C) and the regulating device (RV), and the contour cross section is determined. The distance from the object is determined from the setting of the focusing (S) on the lens (OV). In the microcomputer ( mu C) by means of the camera coordinates and the lens parameters (f,S,B) as well as the distance, in the first instance the centre-of-gravity coordinates of the object are determined and then the object cross section is computed and compared with predeterminable typical intruder cross sections. At the same time, the intruder or his track is monitored over a specified period of time. An alarm criterion is derived from the comparison of the object cross section with possible predeterminable intruder cross sections and from the tracking. <IMAGE>

Inventors

• 👤 Albrecht Dr. Zartner

Bibliographic Details