image-quilting
image-quilting copied to clipboard
trunc8
Implementation of Texture Synthesis and Texture Transfer using Image Quilting. Based on the work of Alexei Efros and William Freeman. Course Project for CS 663 [Digital Image Processing].
Image Quilting for Texture Synthesis & Transfer
Course Project for CS663
Implementation of Texture Synthesis and Texture Transfer using Image Quilting. Based on the work of Alexei Efros and William Freeman.
Usage
First, clone this repository using git clone https://github.com/tezansahu/ImageQuilting.git
CLI Tool:
To use the CLI tool associated with our project do the following:
-
cd code/ - To perform Texture Synthesis use:
python3 main.py --synthesis -i <texture_img> -b <block_size> -o <overlap_size> -t <tolerance> - To perform Texture Transfer use:
python3 main.py --transfer -i1 <texture_img> -i2 <target_img> -b <block_size> -o <overlap_size> -t <tolerance> -a <alpha>
For more details, use python3 main.py -h
TEXTURify Cross-Platform App:
Go into the TEXTURify directory using cd TEXTURify/
To set up the FastAPI Server, do the following:
cd server/
pip3 install -r requirements.txt
uvicorn texturify_api:app --reload
The server would start running on localhost:8000
Now, to set up the Ionic App, spin up a new terminal & type ionic serve [After installing ionic using npm]
This starts the app on localhost:8100.
Disable CORS on your browser (using some extension) and start playing with the app.
Results
Some Results for Texture Synthesis
| Input Image | Synthesized Texture |
|---|---|
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
Some Results for Texture Transfer
| Target Image | Texture Image | Result |
|---|---|---|
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
Created with :heart: by Tezan Sahu, Siddharth Saha & Saavi Yadav
trunc8