Base library input method and device

Abstract

The invention relates to the technical field of image processing, solves the problem that in multiple base libraries, feature comparison results of unlocking processes such as fingerprint unlocking and face unlocking are influenced by methods for storing different feature numbers in each base library at present, and provides a base library input method and device, and the method comprises the following steps: an input step: inputting features into the base libraries; the first judgment step, judging whether the feature number currently input into the bottom library reaches the size of the bottom library or not, if yes, finishing bottom library input, and if not, executing the generation step, and the size of the bottom library being the feature number storable for the bottom library; a generation step: generating false features, and ignoring the false features when the features of the base library are called; and a false feature input step of inputting false features into the base library to execute the first judgment step. By constructing the false features, complementing the base library and completing base library input, the false features directly skip in the process of callingthe base library features, the time required for calling the base library features can be shortened, and the comparison result of the unlocking features and the base library features in the unlockingprocess is not affected.

Description

technical field [0001] The present invention generally relates to the technical field of image processing, and in particular relates to a bottom database entry method, a bottom database feature call method, a feature comparison method and a device. Background technique [0002] Base library entry, that is, the software development kit SDK (Software Development Kit) enters a feature set into the base library, wherein the feature set includes multiple features, and the base library is a storage space that can save features, feature sets or data and / or has storage device. During the unlocking process such as fingerprint unlocking and face unlocking under the screen, users can usually save multiple sets of fingerprint feature sets and face feature sets. The above-mentioned set of fingerprint feature sets and face feature sets are saved to a base library through the SDK. Then multiple sets of fingerprint feature sets and face feature sets are saved to multiple base libraries thr...

AI Technical Summary

Problems solved by technology

[0003] In the multi-bottom library scenario, because the manufacturers supporting the above-mentioned off-screen fingerprint unlocking, face unlocking and other devices do not uniformly stipulate a set of fingerprint and face entry times, the fingerprint feature set, face The number of features in the feature set may be different. During the process of entering features into the base library, since the number of features in each base library may be different, it is impossible to use a unified feature entry method to adapt to the feature entry of multiple bottom libraries.

In addition, with the update of the algorithm and the improvement of the iterative input standard, some of the previously collected atlases may no longer meet the requirements of the new algorithm. In this way, as an algorithm researcher, when training models and algorithm tuning, running images to verify the performance of the algorithm , we have to face the problem of how to deal with such data sets with different bottom library sizes

Images