9/7 LIFT Reconfigurable Architecture Implementation for Image Authentication

Abstract

Considering the information system medical images are the most sensitive and critical types of data. Transferring medical images over the internet requires the use of authentication algorithms that are resistant to attacks. Another aspect is confidentiality for secure storage and transfer of medical images. The proposed study presents an embedding technique to improve the security of medical images. As a part of preprocessing that involves removing the high-frequency components, Gaussian filters are used. To get LL band features CDF9/7 wavelet is employed. In a similar way, for the cover image, the LL band features are obtained. In order to get the 1st level of encryption the technique of alpha blending is used. It combines the LL band features of the secret image and cover images whereas LH, HL, and HH bands are applied to Inverse CDF 9/7. The resulting encrypted image along with the key obtained through LH, HL, and HH bands is transferred. The produced key adds an extra layer of protection, and similarly, the receiver does the reverse action to acquire the original secret image. The PSNR acquired from the suggested technique is compared to PSNR obtained from existing techniques to validate the results. Performance is quantified in terms of PSNR. A Spartan 6 FPGA board is used to synthesize the complete architecture in order to compare hardware consumption.