With the development of artificial intelligence technology, visible light face recognition is becoming more and more popular. But visible light can't solve the problem of facial occlusion. Because the visible light cannot pass through the occlusion and is received by the camera, a lot of information is lost. In reality, when criminals wear masks, they can't recognize faces in visible light. But thermal infrared technology can solve this problem because the heat emitted by the face can be captured by the thermal infrared camera through the mask. In this paper, the generative deep learning method is used to remove the mask in the independent data set, and the recognition result is good.
Face alignment is a research hotspot in face recognition technology, and face alignment can help improve the recognition accuracy of face recognition. Many face alignment methods have been proposed, such as the visible light face alignment technology based on key facial feature points. However, it is not applicable to thermal infrared face alignment technology. This paper proposes a thermal infrared face alignment technology based on regional positioning, that is, the key feature area of the thermal infrared face image is extracted by the generation confrontation network, and then the key feature area is converted into key feature points, so as to realize the thermal infrared face alignment technology. The data set used in this method is the thermal infrared face data set created by the team of Professor Zhang Tianxu of Huazhong University of Science and Technology. The thermal infrared face image after the alignment operation using this method is input to the FaceNet face recognition network, and the recognition accuracy rate is 95.46%. Compared with the input of misaligned thermal infrared face images, the recognition accuracy is improved by 1.62%. The regional positioning thermal infrared face alignment method proposed in this paper can effectively improve the accuracy of thermal infrared face recognition.
With the development of artificial intelligence and deep learning, visible face detection and recognition have always been a research hotspot. However, most visible face detection systems cannot detect and recognize faces without external illumination. Thermal infrared technology is the use of any object higher than absolute zero can emit different wavelengths of electromagnetic radiation, the use of different infrared radiation, an infrared thermal imager can transform the infrared radiation distribution of the object into an image, so it can run in the environment without light. In this paper, the real-time detection of multi-scale and multi-target in the field of view can be realized in the self-collected HUST-MIR Face Database. This method is based on a convolution neural network, which can realize fast and accurate face detection of the thermal infrared image.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.