Final project for pattern recognition subject
Python 3.7 or later with the following pip3 install -U -r requirements.txt
packages:
numpy
torch >= 1.1.0
opencv-python
tqdm
detect.py
runs inference on any sources:
python3 detect.py --source ...
- Image:
--source file.jpg
- Video:
--source file.mp4
- Directory:
--source dir/
- Webcam:
--source 0
- RTSP stream:
--source rtsp://170.93.143.139/rtplive/470011e600ef003a004ee33696235daa
- HTTP stream:
--source http://wmccpinetop.axiscam.net/mjpg/video.mjpg
python detect.py --weights weights/best.pt --cfg cfg/yolov3-tiny-1.cfg --names data/plates.names --source video.mp4 --view-img
I strongly refer:
-
L.P Detection: https://github.com/ultralytics/yolov3
-
Text Detection: https://github.com/clovaai/CRAFT-pytorch
-
Text Recognition: https://github.com/githubharald/SimpleHTR