Language

• English
• 中文

Monitor-DeWatermark

An innovative and efficient watermark removal method designed to accurately identify and elegantly remove watermarks from images

🌄 Framework

├── configs
├── docs
├── modules
│   ├── models
│   │   └── detect
│   │   └── inpainting
│   └── weights
│       └── detect
│       └── inpainting
├── utils
│       └── img_tools.py
├── out
├── test_img
├── 
└── main.py

configs/ : Save the parameters of the model

modules/ : Under the folder is the core code of this project👌

test_img/ : Under the folder are some test images

out/ : Under the folder is the output picture, i. e. the item after the watermark is removed

📝 Introduction

The watermark removal process is divided into two steps:

• The first step is Watermark Detection, with the main code located in the modules/models/detect/ folder, the weights file is saved in the modules/weights/detect/ folder.

• The second step is Watermark Removal, with the main code located in the modules/models/inpainting/ folder. Since the weight file will be updated periodically... No upload.

🤗 Detect weight download (verification code: k1gv)

🤗 Inpaint weight download (verification code: 7urm)

📝 Usage

Show

Demo1

Demo2

Installation

Requires python>=3.8

pip install -r requirements.txt 

Run

Run the project with the following command:

python main.py --input_folder '/path/to/your/input_folder' --output_folder '/path/to/your/output_folder' --config './configs/default.yaml'

• '/path/to/your/input_folder' change the watermark folder ✔️
• '/path/to/your/output_folder' change the output folder🔜