Toward a Secure and Robust Medical Image Watermarking in Untrusted Environment

Soualmi, Abdallah; Alti, Adel; Laouamer, Lamri

doi:10.1007/978-3-319-74690-6_68

Cited by 4 publications

(2 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The work in the field of digital watermarking started with the spatial domain in 1993 [1] and continued in the same domain [4], [8], [9]. Further researchers started working in frequency domain techniques [2], [8], [11] by identifying the advantages of frequency domain techniques over spatial domain in medical field. Later on it is also presented by few researchers that hybrid watermarking techniques are best which incorporates features of multiple digital watermarking techniques combining spatial and frequency / transform domain techniques [3], [5], [8], [31] or combining multiple frequency domain techniques [6], [7], [10].…”

Section: Related Workmentioning

confidence: 99%

“…The method has better transparency with PSNR > 35dB for scaling factor 40. A homomorphic encryption technique using Number Theoretic Transform (NTT) in combination with key exchange protocol Diffie Hellman for security of medical images in cloud is presented by Soualmi et al [2]. Execution time is fast 1.2 -2.8 sec and new transform is tried but robustness is not tested against any attack.…”

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

Hybrid Mode of Medical Image Watermarking To Enhance Robustness and Imperceptibility

Verma¹,

Sharma²

2019

IJITEE

View full text Add to dashboard Cite

In Modern healthcare systems, diagnostic information of patients is managed digitally. Watermarking strategies plays a significant role to ensure the security, authenticity and management of biomedical images and related patient information. This paper explores different invisible Digital watermarking methods both in spatial and transform / frequency domain and then Hybrid method by combining two techniques to justify better results for medical images. Here, host image considered is a medical image and patient’s information is a watermark which is embedded into it. The embedded watermark is hidden in medical host image and its quality should not be degraded to avoid misdiagnosis by doctors. In this paper, LSB, SVD, DCT, DWT and Hybrid techniques (DWT+SVD) are implemented for embedding and extraction process of invisible medical image watermarking. The performance is evaluated based on PSNR as a measure of imperceptibility and CRC & SSIM as a check of robustness. The attacks introduced are cropping, rotating, and noise like Gaussian, Salt & Paper, Speckle and Poisson. Results reveal that DWT provides higher PSNR values but it is robust for few attacks only. On the other side, SVD provides consistent SSIM for all attacks, although PSNR values are less than that of DWT. Hence a hybrid technique is implemented by incorporating benefits of DWT and SVD both and achieved optimal values of PSNR as well as SSIM to provide better security to medical images by preserving the quality as well.

show abstract