The goal of this paper is to detect and recover the tampered regions of the digital image. To achieve this goal, the LSB and the second LSB of the spatial domain was selected to embed the watermark. In particular, the watermark consists of three parts: 1)
recovery code made by the small sized image of the original, 2)
verification code made by the variance value of a block (4 × 4) in the original image (only six MSBs, 1
st ∼ 6
th MSB) and 3)
owner code made by a binary logo image. By using this special watermark, we can detect and recover the tampered regions.
View full abstract