In-vivo detection method based on attitude information
The invention relates to an in-vivo detection method based on attitude information. The method comprises the following steps: (1) detecting a face position on each frame image in an input video; (2) acquiring facial key points and a face shape formed by the key points; (3) performing normalization p...
Saved in:
Main Author | |
---|---|
Format | Patent |
Language | English |
Published |
22.07.2015
|
Subjects | |
Online Access | Get full text |
Cover
Loading…
Summary: | The invention relates to an in-vivo detection method based on attitude information. The method comprises the following steps: (1) detecting a face position on each frame image in an input video; (2) acquiring facial key points and a face shape formed by the key points; (3) performing normalization processing on the face shape based on a 3D face model, and acquiring a rotating angle of the face shape relative to the 3D face model; (4) connecting rotating angles of multiple continuous frame images to form a curve, judging whether the curve meets the set requirement, if so, executing a step (5), otherwise, judging that the face is a false face, and exiting; and (5) judging whether the area of a face color area in a frame image with the largest rotating angle is larger than K, if so, judging that the face is a real face, otherwise, the face is a false face. Compared with the prior art, the method disclosed by the invention has the advantages of low false positive rate, high speed and the like. |
---|---|
Bibliography: | Application Number: CN201510243779 |