Three-dimensional face identification method based on multi-scale covariance descriptor and local sensitive Riemann and sparse classification

Abstract

The invention discloses a three-dimensional face identification method based on a multi-scale covariance descriptor and local sensitive Riemann and sparse classification. The method comprises the following steps of carrying out automatic preprocessing on original G database set face models and P test set face models; according to the database set face models and the test set face models after theautomatic preprocessing in the step (1), establishing a scale space and carrying out multi-scale key point detection and neighbor area extraction; extracting a d*d dimension local covariance descriptor from each key point neighbor area under each scale, and carrying out multi-scale fusion on the local covariance descriptors so as to construct the multi-scale covariance descriptor; and mapping thelocal covariance descriptors to a renewable Hilbert space, and providing the local sensitive Riemann and sparse expression to carry out classification identification on a three-dimensional face. In the invention, a description capability of a single-scale local covariance descriptor can be effectively increased, simultaneously, the local sensitive Riemann and sparse classification can effectivelyuse locality of the multi-scale descriptor.

Description

technical field [0001] The invention relates to the fields of digital image processing and pattern recognition, in particular to a three-dimensional face recognition method based on multi-scale covariance descriptors and local sensitive Riemannian kernel sparse classification. Background technique [0002] Different from two-dimensional images, the three-dimensional face data obtained by the three-dimensional face scanner can effectively contain the inherent spatial geometric information of the face. Because 3D shape data is robust to changes in lighting and views, and unlike 2D data, its pixel values ​​are not easily affected by makeup, etc. These characteristics provide an objective basis for accurate identification of individual identities. With the evolution of the times, the development of anthropometry technology and the enhancement of computing power have greatly promoted the transfer of face recognition methods from purely two-dimensional image-based methods to three...

AI Technical Summary

Problems solved by technology

[0003] However, the face surface changes with age and expression changes, and in practical applications, problems such as complex background and sensor noise make it difficult for a single-scale local covariance descriptor to accurately describe the local features of the face

Images