Title :
Motion estimation and image difference for multi-object tracking
Author_Institution :
IAI, Beer-Yaakov, Israel
Abstract :
This paper presents a multi-object tracking method for video images. This approach can be used for a wide range of applications, such as vehicle tracking for traffic control, security purposes for open/closed areas. This algorithm is based on motion estimation and image difference techniques. It provides an improved capability for object tracking, especially for cases with background conditions due to radiometric changes, camera motion instability and motion of different objects within the images. The method was successfully tested for vehicle traffic control by means of a video camera. Some results will be demonstrated off-line
Keywords :
data compression; image sequences; motion estimation; object recognition; optical tracking; road traffic; video coding; camera motion instability; image difference; motion estimation; multi-object tracking; radiometric changes; security purposes; traffic control; vehicle tracking; video images; Cameras; Image recognition; Motion estimation; Neural networks; Object recognition; Testing; Tracking; Traffic control; Vehicles; Video compression;
Conference_Titel :
Aerospace Conference, 1999. Proceedings. 1999 IEEE
Conference_Location :
Snowmass at Aspen, CO
Print_ISBN :
0-7803-5425-7
DOI :
10.1109/AERO.1999.792106