Acquisition method and detection method of living body face detection head posture regression device
A head posture and face detection technology, applied in the field of image recognition, can solve the problems of non-interactive, unpublished regressor for posture regression, poor anti-attack ability, etc., to improve the accuracy and scientificity, and increase the success rate , the effect of improving speed and accuracy
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Example Embodiment
[0046] Example one
[0047] according to figure 1 , 2 As shown, this embodiment proposes a method for acquiring a head posture regressor for live face detection, including the following steps:
[0048] A. Acquire 2D face model data, 3D face model data, and iris data, and then screen out the acquired 2D face model data, 3D face model data, and iris data to remove the 2D face model data , The low resolution and repeated data in the face 3D model data, remove the damaged data in the iris data;
[0049] B. Select a certain number of feature points in the face two-dimensional model data, and then select the same feature points in the face three-dimensional model data, and map the feature points, and then form the face three-dimensional model data with composite standardization, and then Perform normalization processing on the face three-dimensional model data with compound standardization, and use the distance and direction axis of the two pupils in the face three-dimensional model data ...
Example Embodiment
[0061] Example two
[0062] according to image 3 As shown, this embodiment of a method for detecting a living body face based on a head posture regressor includes the following steps:
[0063] H. Obtain the image of the head posture made by the user according to the instructions issued by the terminal, including two-dimensional images and three-dimensional images, and then use the iris data collected by the iris collection device;
[0064] I. According to the two-dimensional image, three-dimensional image and iris data, the face frame is obtained through the adaboost algorithm;
[0065] G. Locate the coordinates of the face feature points in the face frame by using a supervised gradient descent method;
[0066] K. Perform centralization and normalization of the facial feature points;
[0067] L. According to the processed feature point data, the head angle is obtained through the head posture regressor; when it is judged that the obtained head angle value is within the preset threshold,...
Example Embodiment
[0081] Example three
[0082] according to Figure 4 As shown, the method for detecting a human face in this embodiment includes the following steps:
[0083] M. Obtain the two-dimensional and three-dimensional images of facial expressions made by the user according to terminal instructions, and the iris data collected by the iris collection device;
[0084] N. According to the two-dimensional image, three-dimensional image and iris data, obtain a face frame through an adaboost algorithm;
[0085] O. Locate the coordinates of the face feature points in the face frame by using a supervised gradient descent method;
[0086] P. Perform linear transformation on the located feature points; when it is judged that the feature information value of the feature points after linear transformation is within the preset feature threshold range, the recognition is successful.
[0087] The beneficial effects of the present invention are: by constructing a high-precision head posture regressor, and by a...
PUM
Abstract
Description
Claims
Application Information
- R&D Engineer
- R&D Manager
- IP Professional
- Industry Leading Data Capabilities
- Powerful AI technology
- Patent DNA Extraction
Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.
© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap