Data-Hiding Scheme Using Multidirectional Pixel-Value Differencing on Colour Images
Algorithm 3
The proposed embedding algorithm.
The embedding algorithm is as follows.
Step 1: Divide the colour image into 2 2 size blocks. If the -th block is , then the pixels in are defined as: .
Step 2: Decompose the colour pixels in each block into R, G, and B. is decomposed into ,, and by the following equation: ,,,
Step 3: Regroup the ,, and pixel values in the block to generate ,, and .,,
Step 4: Find the minimum pixel value of , and , respectively. The function returns the minimum value. ,,.
Step 5: Generate pairs in two directions or three directions.Step 5-1: Construct two pairs in two directions based on the minimum value as shown Figure 2. The same algorithm is performed on , and . Therefore, we will explain only . If the minimum pixel value in is , then the two pairs are and .Step 5-2: Construct three pairs in three directions based on the minimum value as shown Figure 3. The same algorithm is performed on ,, and . Therefore, we will explain only . If the minimum pixel value in is , then the three pairs are ,, and .
Step 6: Apply PVD scheme in two directions or three directions.Step 6-1: Perform the PVD scheme to two pairs for embeds the secret data. The two pairs after the PVD scheme is performed which are defined as and .Step 6-2: Perform the PVD scheme to three pairs for embeds the secret data. The three pairs after the PVD scheme is performed which are defined as ,, and .
Step 7: Distribute pixels in two pairs or three pairs to two imagesStep 7-1: Distribute pixels in two pairs to two images and as shown Figure 4.Step 7-2: Distribute pixels in three pairs to two images and as shown Figure 5.
In this step, generated and associated with R channel. and are generated in the G channel, and and are generated in the B channel.
Step 8: Generates two stego colour images and by using the following equation. The function combines R, G, and B to produce a colour image. .
