可復原感興趣區域之醫學影像浮水印技術

ROI Recoverble Digital Watermarking for Medical Images

陳舒菁(Shu-Chin Chen)

指導教授 ： 繆紹綱

中原大學/工學院/電子工程研究所/碩士(2007年)

https://doi.org/10.6840/cycu200700483

摘要

摘要

隨著訊號處理技術及網際網路的快速發展，對多媒體資料的安全(如數位視訊、音樂、影像等資料的未授權傳遞)形成一個極大的挑戰，其中包括保護醫學影像以對抗未授權的存取、公開和竄改。為了能在有安全疑慮的網路上對醫學影像等多媒體資料給予有效的保護，數位浮水印(digital watermarks)技術逐漸受到重視。本論文針對醫學影像提出不會減損診斷品質的新浮水印方法，結合了強健型以及脆弱型浮水印的優點，將醫學影像依照重要性高低區分成兩部份，並分別將兩種不同類型的浮水印依照診斷的重要性嵌入合適的部份來保存診斷品質。嵌入在較不重要區域(NROI)的浮水印使用強健型的嵌入法，其儲存了較重要區域(ROI)之復原資訊，而嵌入較重要區域的浮水印使用脆弱型的嵌入法，將用來做資料完整性的確認，例如偵測局部修改。
我們以濾波、對比度增強、影像壓縮等攻擊模式測試所提方法，並模擬醫學影像由有病徵被竄改為無病徵或者無病徵被竄改為有病徵的情況。實驗結果顯示所提方法能夠確保重要性較低區域(NROI)浮水印的強健性，取出重要性較高區域(ROI)的浮水印後可藉由NROI取出的浮水印復原ROI的原始資訊，且由ROI取回之浮水印的CC (Correlation Coefficient)值可以看出此醫學影像是否遭到竄改，因而此方法對於醫學影像資料的傳遞能增加它的安全性和完整性。

關鍵字：數位浮水印、ROI、醫學影像

關鍵字

ROI ； 數位浮水印 ； 醫學影像

並列摘要

Abstract

As the fast development of signal processing and Internet technology, it is a big challenge to protect multimedia data from the unauthorized spreading of digital video, music, image, etc. The challenge includes protection of medical image data against unauthorized access, disclosure and manipulation. In this thesis, we propose a new watermarking technique that does not degrade diagnostic quality of medical images. The proposed method combines the advantages of both robust and fragile watermarking, where diagnostic quality is preserved by separating a medical image into two regions and embedding two separate watermarks in the regions respectively according to the diagnostic importance of the regions. The original pixel values at the diagnostically more significant region (region of interest, ROI) can be restored using the data contained in the watermark embedded in the less significant region (NROI) by a robust embedding method, whereas the watermark embedded in the diagnostically more significant region is used for data integrity verification, such as the detection of local modification.
In a simulation study, the proposed method is attacked by using filtering, histogram-equalized stretching, image compression, cutting and pasting symptom areas. Experiment results show that the proposed scheme is robust against various attacks for the watermark embedded in NROI because after extracting the watermark in the ROI, the original ROI can be recovered by the recovery information embedded in the NROI. The proposed method is able to identify content modifications in medical images by checking the correlation coefficient of the extracted watermark, thereby providing enhanced security and integrity of distributed medical image data.

Keywords: digital watermark, ROI, medical images

並列關鍵字

medical images ； digital watermark ； ROI

參考文獻

[9] R. C. Gonzalez and R. E. Woods, Digital Image Processing, 2nd ed., Prentice-Hall, 2002.

[11] K. R. Rao and P. Yip, Discrete Cosine Transforms - Algorithms, 2nd ed, Academic Press, 1990.

[12] N. Ahmed, T. Natarajan, and, K. R. Rao, “Discrete Cosine Transform,” IEEE Trans. Computers, vol. C-23, pp. 90-93, Jan. 1974.

[14] D. C. Lou, J. L. Liu, and M. C. Chang, “Digital Watermarking Using Multi-Resolution Wavelet Transform,” Proc. of the IEEE 37th Annual Int. Carnahan Conf. on Security Technology, pp. 370-377, Oct. 2003.

[15] Q. Chen and T. S. Huang, “A DCT-Domain Blind Watermarking Syatem Using Optimum Detection on Laplacian Model,” Proc. of IEEE Int. Conf. on Image Processing, vol. 1, pp. 454-457, Sept. 2000.