An improved reversible hiding method for multimodal biometrics

In this paper, an improved reversible data hiding approach based on correlation analysis for multimodal biometric identification is proposed to ensure the security and integrity of multimodal biometric images. Considering the characteristics of biometrics, correlation analysis is first utilized between the face image and the palm image to enhance the security, and a residual image is obtained by exploiting the spatial and content correlation. Then an improved method is presented to embed the residual palm image into face image for secure transmission. Finally, image authentication is performed for further multimodal biometric identification at the receiver. Once images are authentic, identification is carried out for further identity authentication. Experimental results demonstrate that our proposed method has better quality than the previous work and the perfect security performance is validated on biometric identification.