全部 标题 作者
关键词 摘要

OALib Journal期刊
ISSN: 2333-9721
费用:99美元

查看量下载量

相关文章

更多...

A Medical Image Watermarking Technique for Embedding EPR and Its Quality Assessment Using No-Reference Metrics

Keywords: No Reference , Quality Measurement , EPR (Electronic Patient Record) , Mean Square Error (MSE) , Peak Signal to Noise Ratio (PSNR)

Full-Text   Cite this paper   Add to My Lib

Abstract:

Digital watermarking can be used as an important tool for the security and copyright protection of digital multimedia content. The present paper explores its applications as a quality indicator of a watermarked medical image when subjected to intentional (noise, cropping, alteration) or unintentional (compression, transmission or filtering) operations. The watermark also carries EPR data along with a binary mark (used for quality assessment). The binary mark is used as a No-Reference (NR) quality metrics that blindly estimates the quality of an image without the need of original image. It is a semi-fragile watermark which degrades at around the same rate as the original image and thus gives an indication of the quality degradation of the host image at the receiving end. In the proposed method, the original image is divided into two parts- ROI and non-ROI. ROI is an area that contains diagnostically important information and must be processed without any distortion. The binary mark and EPR are embedded into the DCT domain of Non-ROI. Embedding EPR within a medical image reduces storage and transmission overheads and no additional file has to be sent along with an image. The watermark (binary mark and EPR) is extracted from non-ROI part at the receiving end and a measure of degradation of binary mark is used to estimate the quality of the original image. The performance of the proposed method is evaluated by calculating MSE and PSNR of original and extracted mark.

Full-Text

Contact Us

service@oalib.com

QQ:3279437679

WhatsApp +8615387084133