Facial expression recognition method and applied intelligent lock system thereof

Abstract

The invention discloses an anti-intrusion intelligent lock system based on facial expression recognition. The method according to the anti-intrusion intelligent lock system comprises the following steps that a, facial image information is acquired from video information; b, facial expression feature points in the facial image information are extracted based on a shape model; c, motion tracking of the facial expression feature points matched with the DWT-SIFT features in the video information is performed by using the circular neighborhood based on the DWT-SIFT features so as to obtain the motion feature information of the facial expression feature points; d, classified recognition of the motion feature information is performed through a classifier so as to obtain expression state data. The processing module in the intelligent lock system can obtain the expression of the user through the expression state data, and the processing module compares the expression of the user and preset expression data through the preset expression data so as to control operation of the intelligent lock. The user is enabled to control operation of the intelligent lock through the facial expression in danger and perform corresponding alarming so that the intelligent lock system is enabled to be safer and the user can perform operation under protection.

Description

technical field [0001] The invention relates to image processing technology, in particular to a facial expression recognition method and an intelligent lock system for its application. Background technique [0002] Based on face recognition technology, it has been widely used in daily life. Computer vision technology is used to identify facial feature information, including collecting face images and tracking face images, extracting face recognition features, and learning and memorizing face recognition data. and comparative identification. Internationally, face recognition research started earlier than domestically, and the technology is more advanced than domestically. Well-known research institutions include MIT in the United States and Department of Engineering in University of Cambridge in the United Kingdom. After the development of domestic face recognition technology in recent years, the face recognition rate under ideal lighting, facial expression and posture cond...

AI Technical Summary

Problems solved by technology

The above methods have improved the robustness of face recognition in some aspects, but the universality is not strong, and the recognition performance of expression changes is average.

[0004] Early facial expression recognition was mainly based on static images. This recognition method is simple in operation and fast in speed, but it requires the image to be processed to have obvious or exaggerated expression features, and it is difficult to recognize subtle facial expressions. The robustness is poor, and the recognition rate is low. lower

[0005] The existing face recognition technology applied to the smart lock system, and the traditional smart lock system using face recognition, all recognize the static expression of the face, such as security in banks, gold stores, vaults, art warehouses, etc. In key industries, when the staff encounters threats, they cannot detect danger in time or call the police remotely. The facial expression is a signal that the staff can easily control, so we hope to develop the corresponding facial expression recognition method to use with the smart lock system

Images