Face information automatic login method based on face recognition technology

Abstract

The invention discloses a face information automatic login method, adopting the face detection and recognition technology and automatically logging on based on the face information with high frequencyextracted from the collected video data or multiple photographs by the automatic login system. The method comprises following steps: (1) adopting the face detection technology, detecting the face from the data source, wherein the data source can be video data collected by a pick-up head or pictures; (2) off-line training step, adopting the face recognition technology to extract the detected facecharacteristic value, adding in the characteristic value data-base; adopting the face recognition technology to match the face in the characteristic value data-base, finding out the face information with high similarity in the data-base, considering as the face information of the same people; (3) on-line constraining step, under the constraining condition such as time and space, judging the off-line train result, combining the off-line train and the on-line constraining, accurately judging the face information of the same people in the data-base; (4) finding out the face information with the highest frequency in the data-base, and automatically logging on the data-base.

Description

Technical field: [0001] The invention relates to a face information automatic registration method based on face recognition technology, which belongs to the technical field of face recognition. Background technique: [0002] At present, due to its intuitiveness, convenience, non-contact, friendliness, and high user acceptance, the face recognition system has been widely used in many aspects such as enterprise management, residential security, and identity authentication. [0003] The working steps of the face recognition system are generally divided into two parts: face information registration and face matching query. Usually, in the face information registration step, use the camera to capture the face information of the target object, or enter a photo of the target object, use face recognition technology to extract the feature value of the face information, and log in the face information database for Matching query is used; in the face matching query step, the currently...

AI Technical Summary

Problems solved by technology

Studies have shown that the feature difference of the same person under different lights is even greater than the feature difference between different people under the same light. The ambient light has a great impact on the recognition performance. When matching queries, the ambient light is different from the login , the accuracy of the query will be greatly reduced, which cannot meet the requirements of the application

[0007] (3) Due to the one-to-one entry, there is only one piece of data in the database for each face information, and the feature values ​​are not comprehensive enough, which will also reduce the accuracy of recognition matching queries

This traditional login method cannot be used for some smart devices without user input interfaces, such as robots, so it is not conducive to generalization and expansion

Images