Volume XLII-2/W12
Int. Arch. Photogramm. Remote Sens. Spatial Inf. Sci., XLII-2/W12, 61-67, 2019
https://doi.org/10.5194/isprs-archives-XLII-2-W12-61-2019
© Author(s) 2019. This work is distributed under
the Creative Commons Attribution 4.0 License.
Int. Arch. Photogramm. Remote Sens. Spatial Inf. Sci., XLII-2/W12, 61-67, 2019
https://doi.org/10.5194/isprs-archives-XLII-2-W12-61-2019
© Author(s) 2019. This work is distributed under
the Creative Commons Attribution 4.0 License.

  09 May 2019

09 May 2019

DIGITAL WATERMARKING OF 3D MEDICAL VISUAL OBJECTS

M. N. Favorskaya and E. I. Savchina M. N. Favorskaya and E. I. Savchina
  • Reshetnev Siberian State University of Science and Technology, Institute of Informatics and Telecommunications, 31, Krasnoyarsky Rabochy ave., Krasnoyarsk, 660037 Russian Federation

Keywords: Digital Watermarking, 3D Medical Object, Digital Wavelet Transform, Digital Hadamard Transform

Abstract. At present, medical equipment provides often 3D models of scanning organs instead of ordinary 2D images. This concept is supported by Digital Imaging and COmmunications in Medicine (DICOM) standard available for telemedicine. This means that the confidential information under transmission ought to be protected by special techniques, particularly digital watermarking scheme instead of textual informative files represented, for example, on CD disks. We propose a multilevel protection, for which a fragile watermark is the first level of protection. The Region Of Interest (ROI) watermark and textual watermarks with information about patient and study (the last ones can be combines as a single textual watermark) form the second level of protection. Encryption of the ROI and textual watermarks using Arnold’s transform is the third level of protection. In the case of 3D models, we find the ROI in each of 2D sliced images, apply the digital wavelet transform or digital shearlet transform (depending on the volume of watermarks) for the ROI and textual watermarks embedding, and embed a fragile watermark using digital Hadamard transform. The main task is to find the relevant regions for embedding. To this and, we develop the original algorithm for selecting relevant regions. The obtained results confirm the robustness of our approach for rotation, scaling, translation, and JPEG attacks.