| 摘要 |
PROBLEM TO BE SOLVED: To provide a video processor that enables a user to easily recognize how many object, such as intruders, are present within a prescribed area. SOLUTION: Area extract sections 114, 116, etc., extract area data 120, 122, etc., of intruders on the basis of digital video images 108, 110 photographed by cameras 101, 104, etc., and body height measurement sections 2, 14, etc., detect body height data 4, 16, etc. Area normalization sections 6, 18, etc., normalize the data, change in top of head detection sections 26, 28, etc., obtain a temporal change in tops of heads on the basis of the normalized data, and a correlation calculation section 34 and a video classification section 38 classify the same motion of the top of heads into the same group. A video image selections section 40 selects those facing front as object video images of the intruders and displays them on a monitor.
|
