Face recognition method based on intelligent security and protection robot under unlimited condition

Abstract

The invention discloses a face recognition method based on an intelligent security and protection robot under an unlimited condition, and the method comprises the steps: carrying out the face detection of an image through employing the HOG features, and extracting a face region image for the subsequent processing; secondly, training a classifier for performing binary classification on the clear image and the blurred image by utilizing the CNN convolutional neural network, performing deblurring processing on the classified blurred image, and skipping the step of the clear image; then, carryingout deblurring processing on the moving image frame sequence by using a Lucy-Richardson algorithm; normalizing the image to the same size scale by using an SRCNN image super-resolution algorithm; inputting the image processed by the above steps into a trained CNN convolutional neural network to extract a feature vector; and finally, classifying the extracted feature vectors through an SVM, therebyrealizing face recognition. The method has the advantage that the efficiency and precision of face recognition of the intelligent security and protection robot in the patrol process are effectively improved.

Description

technical field [0001] The invention relates to the technical field of computer vision, and more specifically, to a face recognition method based on an intelligent security robot under unrestricted conditions. Background technique [0002] Existing security mainly uses video surveillance, which requires dedicated security personnel to check 24 hours a day and security personnel to perform patrol tasks regularly. The human resources are expensive, and monitoring and patrolling by people is prone to omissions at night when they are tired. Condition. And when a large-scale power outage occurs, most of the traditional security systems will be paralyzed. Intelligent security robots can be applied to indoor and outdoor security patrol tasks, and collect environmental data to form a flexible monitoring system. Security robots can continue security patrols in relatively harsh environments, greatly saving human resources, and have absolute advantages in safety. Since the image qua...

AI Technical Summary

Problems solved by technology

[0003] In view of this, the present invention proposes a face recognition method based on an intelligent security robot under unrestricted conditions. The method solves the problem of the existing intelligent security robot patrolling through image deblurring, light changing, and super-resolution image frame sequence processing. In the process, the problems of reduced face recognition accuracy or unsuccessful recognition caused by blurred images, complex and changeable lighting conditions, changes in face poses, and low resolution of face images

Images