Secure visual cryptography for medical image using modified cuckoo search

Abstract

Optimal secure visual cryptography for brain MRI medical image is proposed in this paper. Initially, the brain MRI images are selected and then discrete wavelet transform is applied to the brain MRI image for partitioning the image into blocks. Then Gaussian based cuckoo search algorithm is utilized to select the optimal position for every block. Next the proposed technique creates the dual shares from the secret image. Then the secret shares are embedded in the corresponding positions of the blocks. After embedding, the extraction operation is carried out. Here visual cryptographic design is used for the purpose of image authentication and verification. The extracted secret image has dual shares, based on that the receiver views the input image. The authentication and verification of medical image are assisted with the help of target database. All the secret images are registered previously in the target database. The performance of the proposed method is estimated by Peak Signal to Noise Ratio (PSNR), Mean square error (MSE) and normalized correlation. The implementation is done by MATLAB platform.