In this repos we study number plate detection and recognition using different deep learning models and computer vision approches.

In order to detect licence we will use Yolo ( You Only Look Once ) deep learning object detection architecture based on convolution neural networks. This architecture was introduced by Joseph Redmon , Ali Farhadi, Ross Girshick and Santosh Divvala first version in 2015 and later version 2 and 3.

Yolo v1 : Paper link.

Yolo v2 : Paper link.

Yolo v3 : Paper link.

Yolo is a single network trained end to end to perform a regression task predicting both object bounding box and object class. This network is extremely fast, it processes images in real-time at 45 frames per second. A smaller version of the network, tiny YOLO, processes an astounding 155 frames per second.

You will find more information about how to train Yolo on your customized dataset in this Link.

There is also other Deep learning object detector that you can use such as Single Shot Detector (SSD) and Faster RCNN.

We used python v3.5.5 install requirement

pip install -r requirement.txt

Download Yolo weights from this Link.

Detect LP from an image

python detector.py --image test.jpg

To detect LP from a video

python detector.py --video test.mp4

Detection from image :

We are stadying Tunisian plates and USA plates for the recognition, check the sub folders in plates recognition folder!