A service that presents the optimal laundry course by detecting laundry contamination using YOLOv3

2021-1 Capstone Design2

2021.03 - 2021.06

📖 Contents

1. Darknet
2. Dataset
3. Config
4. Train
5. Web

---

1. Darknet

• AlexeyAB/darknet

  mAP, loss, iteration 등을 추적하면서 연구하기 편해서 사용했다.

2. Dataset

• Google Open Images Dataset V6(Clothes, Button data)
• Tilda Textile Texture-Database(Stain data)
• Kaggle(Stain data)
• Google Image(Stain, Button data)

• Yolo Mark

  Marking bounded boxes of objects in images for training Yolo v3

3. Config

📂 Darknet
┣ yolov3-spp_custom.cfg
┃
┣ obj.data
┃ ┣ train.txt
┃ ┗ valid.txt
┃
┗ obj.names

4. Train

📂 jupyterlab
┗ YOLO_MODEL.ipynb

5. Web

python detect.py

---

Reference

• AlexeyAB/darknet