rembg

concept

---

U2-NET이라는 Object Detection Deep Learning model을 이용해 배경 이미지를 날리고 테두리 Alpha 부분을 다듬어서 깔끔하게 배경을 날려주는 네트워크.

code

use web

from flask import Flask, render_template

app = Flask(__name__)

@app.route('/', methods=['GET', 'POST'])
def index():
  return render_template('index.html')

if __name__ == '__main__':
  app.run(debug=True, port=5001)

## method Segmenation + Removal 으로 background removal

Segmentation

The algorithm calculates foreground edges and segments it out.

Removal

참조

• https://github.com/danielgatis/rembg
• https://www.youtube.com/watch?v=r29j8u7XwoM