@techreport{oai:ipsj.ixsq.nii.ac.jp:00053774, author = {川端, 敦 and 谷藤, 真也 and 諸岡, 泰男 and Atsushi, Kawabata and Shinya, Tanifuji and Yasuo, Morooka}, issue = {62(1984-CVIM-034)}, month = {Jan}, note = {This paper deals with a new method to extract a moving object from time varing image based on a human vision recognition model. At the first all edges of an input image are enhanced by Hitachi's Image Signal Processor. A background image which does not move or moves very slowly is gradualy constructed into a background memory. The difference between the input image (edge enhanced image) and the background image (edge enhanced image) is calculated in each pixels. After that the image of a moving objects is extracted according to the value of this difference., This paper deals with a new method to extract a moving object from time varing image, based on a human vision recognition model. At the first, all edges of an input image are enhanced by Hitachi's Image Signal Processor. A background image which does not move or moves very slowly is gradualy constructed into a background memory. The difference between the input image (edge enhanced image) and the background image (edge enhanced image) is calculated in each pixels. After that, the image of a moving objects is extracted according to the value of this difference.}, title = {移動物体像の抽出技術}, year = {1985} }