Infrared photography-based bio-feature 3D four-dimensional data identification method and system

Abstract

The invention provides an infrared photography-based bio-feature 3D four-dimensional data identification method and system. The method comprises the steps of preprocessing 2D color images collected for a target organism by one or more zoom color cameras and an infrared image collected for the target organism by a depth infrared camera to obtain spatial point cloud information of the target organism in a given time, performing registration and fusion to obtain 3D data of the target organism in the given time, and finishing model rebuilding; forming a database by taking identity information of the organism as an identification mark; and finding bio-feature 3D four-dimensional data stored in the database by utilizing the identity information of the target organism, and performing corresponding point cloud comparison to identify an identity of the target organism. According to the method and the system, the 2D high-definition color images of the color cameras and depth 3D information collected by the depth infrared camera are subjected to the registration and fusion, so that the extraction of feature points is reduced, the complexity of an algorithm is lowered, and the rebuilding precision and efficiency are improved.

Description

technical field [0001] The invention relates to the technical field of biological feature identification, in particular to a method and system for identifying biological feature 3D four-dimensional data based on infrared photography. Background technique [0002] Biological characteristics are the inherent physiological or behavioral characteristics of living things, such as fingerprints, palm prints, irises or human faces. Biological characteristics have certain uniqueness and stability, that is, the difference between any two biological characteristics is relatively large, and biological characteristics generally do not change greatly over time, which makes biological characteristics very suitable for application In scenarios such as identity authentication or authentication information in the identification system. [0003] In the current binocular scheme, the registration method of basic grayscale, domain change and basic features is generally used. Firstly, feature ext...

AI Technical Summary

Problems solved by technology

[0004] 1) When the feature points in the image are not obvious or the resolution is low, it is difficult to extract feature points from the image

[0005] 2) When the scene is similar, the coordinate correspondence between the two images is not linear, and the use of nonlinear transformation has great limitations

[0006] 3) To find feature points from the image and further perform feature matching, a large number of complex algorithms are required in the whole process, and the higher the resolution, the higher the precision required, the more complex the calculation, therefore, high computing performance is required. In addition The final calculation result is not stable

Images