Vietnamese License Plate Recognition

Tasks

• Detect license plate(s) in an image, a video or realtime.
• Extract characters from detected license plate(s).

How to inference

First of all, make sure that you have installed all requirements. Then, to inference, you can run the Python file with the following code

python detect.py \
        --modality <image-or-video-or-realtime> \
        --confidence <confidence-threshold-for-ocr> \
        --src <path-to-videos-or-images> \
        --out <path-to-output-directory>

Or the provided script files

# For image
bash detect_image.sh

# For video
bash detect_video.sh

# For realtime
bash detect_realtime.sh

Results

Challenges

Cannot detect or OCR on license plates in the environment that:

• Too bright
• Too blurry
• License plates with characters that are not black (because the OCR model was not trained for these cases)
• License plates that are too small

Refereces

https://github.com/trungdinh22/License-Plate-Recognition https://github.com/WaseemKn/FuzzyColorImageEnhancement-FuzzyLogicCourse-ITE5thYear