Abstract
There are still many problems in the application of face recognition technology in specific environments. In order to improve the recognition accuracy of face recognition technology in large-scale event scenes, a face tracking algorithm is constructed based on improved ORB features in this paper. The algorithm uses the kernel function color histogram to model the target, which can perform well on target recognition during rotation or edge occlusion. The value of the face representation function is calculated by combining the texture feature and the color feature. By converting the function value into moment feature and combining it with the Bhattacharyya coefficient, the tracking performance of the target and the accurate detection of the target real scale improves. Taking the Tianjin National Games as an example, the intelligent security system based on face recognition was designed and implemented. The test results show that the proposed algorithm can reduce the face feature dimension effectively. When the control library is no more than 10,000 people, the false alarm rate is 1%, and the false negative rate is less than 5%. It provides a new recognition method for face recognition in the application of large-scale event scenes.
Get full access to this article
View all access options for this article.
