Task B of 2023 Shenzhen Cup Mathematical Contest in Modeling
This project aims at adding watermark to pictures without being noticed by viewers.
Task B1
python DCT_vote.py --img P.jpg --logo 深圳杯数学建模挑战赛 --output SP.png
python DCT_vote_inv.py --img SP.png
The output should be:
Watermark: 深圳杯数学建模挑战赛
Task B2
python DCT_vote.py --img P.jpg --text law.txt --output SP.png
python DCT_vote_inv.py --img SP.png
Then you can see an identical text file watermark.txt in ./output/.
Images of .jpg and .jpeg formats are available now!
Updated on August 9th, 2023. This version uses DCT to hide watermark in picture. Technical details will be shown in the paper.
Configurations
See at version 0.0.2.
Commands
- For adding watermark:
python DCT_vote.py --img A --logo B --text C.txt --output D
Check out the meanings of arguments at version 0.0.2.
- For extracting watermark:
python DCT_vote_inv.py --img D
Will extract watermark added by DCT_vote.py in picture ./output/D. If the watermark's length exceeds 100, then check it out at ./output/watermark.txt.
Images of JPEG format is not supported in this version.
Note that the name of argument has changed compared to version 0.0.2.
Updated on August 4th, 2023. This version includes the simplest way of hiding watermark (a.k.a. steganography), LSB algorithm.
Configurations
All input files should be placed under the input directory, including the initial picture and text you want to add as watermark (optional).
Commands
- For adding watermark:
python SimpleLBS.py --img A --logo B --text C.txt --output D
Where A is the filename of the input picture and D is the filename of the output picture (output/D will appear after running this command).
If you want to add a long text as watermark, put the whole text in a text file named C.txt. Otherwise, you can input your watermark as argument B.
All arguments are optional.
- For extracting watermark:
python SimpleLBS_inv.py --output D
Will extract watermark added by SimpleLBS.py in picture ./output/D.