Muhammad F. Islam and Mohammed N. Islam
biometric features, color image, orthogonal coding, pseudo-random numbers, Walsh code
A novel approach is proposed in this paper for securing multiple binary biometric signatures by encrypting, multiplexing and then embedding them into a color image. Multiple binary biometric signatures, such as fingerprint, iris scan, signature, and personal identification text information, are encrypted individually using orthogonal codes and then multiplexed together to form a single two dimensional binary image matrix. A color cover image is then decomposed into three color channel images, namely, red, green and blue. The matrix is then embedded into one of the channel images using the random bit selection process based on user-supplied keys, which chooses one from three least significant bits of the cover image to hide the bit of the biometric information. Three sets of encrypted binary information are embedded into three color component images. Then all three stego images are combined to form the color stego image, which hides a number of biometric information inside the image and makes them invisible to intruders. Computer simulation results using sample color images and biometric information show no visual distortion in stego image as compared to cover image. Also the histogram analyses of cover and stego images confirm that there are no significant changes in the image because of these embedding processes. The proposed embedding process is highly secure such that no information can be retrieved without knowing the process and the correct keys used for encryption. Orthogonal coding technique further enhances the security by increasing robustness and making it practically impossible to access the information without authorization. Simulation experiments prove that most steganalysis methods would fail in detecting any meaningful information from the stego image.
Important Links:
Go Back
