| Input: All perceptual hash feature vectors , |
| , where |
| and , which are the output |
| of Algorithm 1. |
| Output: A map that includes the detection results. |
| Step 1. Creating packages, denoted as , |
| , and , where |
| and is a preset threshold. |
| Step 2. All perceptual hash feature vectors |
| are stored into the |
| packages, respectively, according to |
| the value of . |
| Step 3. A map is created with the same size of |
| suspicious image and all its initial pixel |
| values are set to zero. |
| Step 4. For each package |
| Step 5. The block pairs contained in will be |
| matched according to their perceptual |
| hash feature vectors and coordinate |
| positions. The values of the corresponding |
| coordinate positions in the map will be |
| set to a same pixel value “255” according |
| to the coordinates of the suspicious |
| image if the block pairs are diagnosed |
| as similar. |
| Step 6. For each block contained in package , |
| it will be matched with all blocks |
| contained in package if |
| with the same method |
| of Step 5. |
| Step 7. End For |
| Step 8. Outputting the map. |
