In developing countries such as India, transmission of medical images over the network is essential for providing health services to rural/remote population. During this transmission, a small distortion in the medical image may lead to wrong conclusions. Thus, arises the need to embed some authentication information in medical images to authenticate transmitted image. In this study, a blind and fragile watermarking technique based on Weber's law is proposed. Here, fragile watermark is inserted by quantising intensity of selected pixels in a medical image. As the proposed technique is fragile, it detects even a small change made to watermarked image and locates blocks where distortion has been made. Besides being able to detect all malicious changes that affect the quality of watermarked image, it is tolerant to image compression, which is essential to transmit images over the network. The proposed technique is highly imperceptible as the watermark is inserted by modifying intensity of dark pixels selected using Weber's law. An extension of the proposed technique is also given to enhance its capacity. Experimental results on standard medical image database prove the efficacy of the proposed method over existing works.

No abstract

Digital Watermarking allows an individual to add some hidden copyright notices or other verification messages to digital media, where message is a group of bits describing the information pertaining to the signal or its author. For this purpose, two techniques viz. global watermarking and local watermarking can be used. In this paper, the performance of two feature based global and local watermarking techniques using Zernike Moments(ZMs) have been analyzed by evaluating their robustness against geometric, photometric and other signal processing attacks including rotation, cropping and noise. Experimental results have been provided in order to compare ZMs based global and local watermarking techniques for different types of attacks. Recommendations have been made based on the comparison of these techniques with other existing works.

Zernike Moments (ZMs) are used in many image processing applications, due to their resistance against various signal processing and geometric attacks. Digital image watermarking is one of those application areas, where ZMs are widely used to insert and extract the watermark bits for digital media authentication. In all the existing ZM based watermarking techniques, magnitude of moments is used to insert and extract the watermark. In this paper, the authors’ have proposed a semi blind watermarking technique in which phase of ZMs is used for watermark insertion and extraction. Due to the use of phase of ZMs, 100% detection ratio is achieved against any geometric and other signal processing attacks. To make the proposed technique fast, q-recursive method is used to compute the Zernike polynomials. The use of q-recursive method has also increased the transparency of watermark due to its better reconstruction ability as compared to traditional moment computation method. Through detailed experimentation, it has been confirmed that the proposed watermarking technique is fast, has more imperceptibility, less Bit Error Rate (BER) and more capacity as compared to traditional ZMs magnitude based watermarking technique.

No abstract