Digital Image Processing

N. Venkatram ¹, L. S. S. Reddy ², P. V. V. Kishore ³

Affiliations
1 KL University, ECM Dept., Green Fields, Vaddeswaram, Guntur DT, IN
2 KL University, CSE Department, IN
3 ECE Department, KL University, IN

Abstract

Medical image watermarking (MIW) has challenged researchers for over a decade now. Medical images are difficult to watermark as they contain sensitive information of a patient’s disease. During transportation through internet the medical images are subjected to various attacks. For telemedicine applications it becomes vital to prevent the authenticity regarding patient’s records. A two dimensional discrete wavelet transform based approach is designed and tested to embed a patient passport photograph into his or her medical image. This method computes 2D DWT at various levels and mixes the approximate coefficients using a wave coupling coefficient. The proposed watermarking algorithm is tested for available mother wavelets and around five levels of decomposition with multiple coupling coefficients. Peak Signal-to-Noise ratio, Minimum Mean Square Error and Normalized cross correlation coefficient are computed to identify the best wavelet and the level of decomposition that can significantly be applied to medical image watermarking. The results demonstrate that db2 at level 4 is the top performing wavelet.

Keywords

Medical Image Watermarking, 2D Discrete Wavelet Transform, Wave Coupling Coefficient, Peak Signal-To-Noise Ratio, and Minimum Mean Square Error.

Full Text