Invention Grant
- Patent Title: Facial recognition system, facial recognition server, and facial recognition method
-
Application No.: US15578920Application Date: 2016-05-27
-
Publication No.: US10592731B2Publication Date: 2020-03-17
- Inventor: Masashige Tsuneno , Kaoru Tsurumi , Norio Saitou
- Applicant: Panasonic Intellectual Property Management Co., Ltd.
- Applicant Address: JP Fukuoka
- Assignee: PANASONIC I-PRO SENSING SOLUTIONS CO., LTD.
- Current Assignee: PANASONIC I-PRO SENSING SOLUTIONS CO., LTD.
- Current Assignee Address: JP Fukuoka
- Agency: Wenderoth, Lind & Ponack, L.L.P.
- Priority: JP2015-120455 20150615
- International Application: PCT/JP2016/002567 WO 20160527
- International Announcement: WO2016/203717 WO 20161222
- Main IPC: G06K9/00
- IPC: G06K9/00 ; H04N7/18

Abstract:
A statistical processing is performed without performing complicated determination, so as not to repetitively include the same person. A facial feature extractor (32) in a facial recognition server (30) extracts features of facial image data including a face which is shown in an image obtained by a camera device (10) imaging an imaging area. A preservation unit (35) preserves features of facial image data including a face which has passed through the imaging area, in a pre-passing comparative original facial feature data memory (41). A statistics unit (37) performs statistical processing on the features of the facial image data, which are extracted by the facial feature extractor (32), in a case where feature having high similarity which is obtained by comparison with the features (extracted by the facial feature extractor (32)) of the facial image data imaged by the camera device (10) and is equal to or greater than a predetermined value are not preserved in the pre-passing comparative original facial feature data memory (41) until a predetermined time from an imaging time point by the camera device (10).
Public/Granted literature
- US20180157898A1 FACIAL RECOGNITION SYSTEM, FACIAL RECOGNITION SERVER, AND FACIAL RECOGNITION METHOD Public/Granted day:2018-06-07
Information query