Forgery Detection of Images and Videos based on Image Watermarking and Image Matting

• Main Authors: Chen, Wei-Hao, 陳偉昊
• Other Authors: Hu, Wu-Chih
• Format: Others
• Language: zh-TW
• Published: 2012
• Online Access: http://ndltd.ncl.edu.tw/handle/04582106277403860234

碩士 === 國立澎湖科技大學 === 電資研究所 === 100 === This thesis proposes a forgery detection of images and videos based on image watermarking and image matting. In this thesis the “Robust image watermarking scheme based on DWT+DCT+SVD” is proposed. The cover image is transformed from the RGB color space into YCbCr color space to obtain a grey-level image. The DWT is then applied to the grey-level image to obtain LL-subband image. Next, DWT is applied to the LL-subband to obtain the frequency components. Finally, SVD is applied to the frequency components to embed the watermark. In the proposed forgery detection, the singular value matrix of the proposed image watermarking and the alpha matte of image matting are used to obtain the forgery detection of foreground image and background image. The difference of singular value matrix is used to determine the image was forged or exchanged to obtain the forgery detection of images. Experimental results show that the proposed robust image watermarking scheme based on DWT+DCT+SVD outperforms state-of-the-art image watermarking algorithm. Furthermore, the proposed forgery detection based on image watermarking and image matting can obtain correct detection under various forged images and exchanged images. Therefore, the proposed method in this thesis has a high value on the forgery detection of images and videos.