Informatics Publishing Limited

Rohit M. Thanki ¹, Rahul K. Kher ², Divyang D. Vyas ³

Affiliations
1 EC Department, Government Polytechnic, Rajkot, Gujarat, IN
2 EC Department, G H Patel College of Engineering & Technology, Vallabh Vidyanagar, Gujarat, IN
3 EC Department, Darshan Institute of Technology & Engineering, Rajkot, Gujarat, IN

Abstract

Because of the importance of the security issues in the management of medical information on internet and provide copyright protection to medical images, we suggest using a watermarking scheme in transform domain which may help medical data protection and authentication. So Digital Watermarking for medical images has recently been explored as it can be used to hide the patient's information and then extract back the same. In this paper, we have implemented and analysed two transform domain watermarking techniques viz. Frequency domain technique for different Mid-Band DCT coefficients and Wavelet domain technique for different noise values have been implemented for medical images likes CT, MRI and Ultrasound and check detector performance of both techniques. The detector performance for both the techniques is quite encouraging, which is indicated from the PSNR and NCC values of above 63 dB and 0.9, respectively for both techniques. However, the results show that the wavelet domain technique performs better than the frequency domain technique in terms of recovery of watermarks.

Keywords

Discrete Cosine Transform (DCT), Discrete Wavelet Transform (DWT), Digital Watermark, Medical Images, PSNR, NCC.

Full Text

     

Rohit M. Thanki ¹, Rahul K. Kher ², Divyang D. Vyas ³

Affiliations
1 EC Department, Government Polytechnic, Rajkot, Gujarat, IN
2 EC Department, G H Patel College of Engineering & Technology, Vallabh Vidyanagar, Gujarat, IN
3 EC Department, Darshan Institute of Technology & Engineering, Rajkot, Gujarat, IN

Abstract

In this paper two digital watermarking schemes for embedding a monochromic watermark image into color cover image for different noise sequence like Pseudorandom (PN) and White Gaussian Noise (WGN) have been implemented. In order to exploit the correlation properties of both PN and WGN, the correlation based watermarking techniques like-Threshold Based Detection and Comparison Based Detection in Spatial domain have been analyzed for different noise power values and block size. Although the watermarked images appear quite different, visibly from the original image after noise sequence is embedded. As a result of our work, invisible watermarked image is generated by using PN Sequence and visible watermarked image is generated by using WGN Sequence for high value of noise. This paper also gives comparison of these schemes for different noise power values and watermarked images have been verified on the parameters of PSNR (Peak Signal to Noise Ratio) and MSE (Mean Square Error).

Keywords

Pseudorandom Noise (PN), White Gaussian Noise (WGN), Processing Time of Embedder (PTE), Processing Time of Detector (PTD), Threshold Value.

Full Text