In this project we demonstrate how to use OpenCV only, with traditional computer vision approaches, to perform License Plate Detection (LPD).
We follow two approaches:
1- Morphology based approach: where only morphological transforms are used, along with some rules to detect the LP.
2- Charater based approach: in addition to basic morphological approaches, basic char detection, also based on morphology, is used as an extra characteristic of the LP.
Further, the problem of Licence Plate Recognition (LPR), by recognizing the number and digits written, can be addressed by the second approach.
In both approaches, we load HD videos (1080p). Due to the camera position, this is the most effective resolution to detect LP patterns.
In both approaches we merge car detection, using background subtraction, to narrow the search space.
For more details, please see full documentation, and the executable example notebook
You need to install the packages in requirements.txt:
pip install -r requirements.txt
You can run over a video, or single image as follows:
You can call the process_video as follows:
from video import process_video
from char import detect_LP
video_file = 'dat/detection_test.mp4'
video_output_file = 'dat/char_LP_detection.mp4'
process_video(video_file, video_output_file, detect_LP_fn=detect_LP)100%|██████████| 2171/2171 [04:54<00:00, 7.36it/s]
Video is ready at: dat/char_LP_detection.mp4
Note that, you can choose the approach by importing detect_LP from char or morpho
You can also do the same from the command line:
!python main.py --video_file dat/detection_test.mp4 --video_output_file dat/cars_detection.mp4 --detect_LP_fn 1OpenCV: FFMPEG: tag 0x47504a4d/'MJPG' is not supported with codec id 7 and format 'mp4 / MP4 (MPEG-4 Part 14)'
OpenCV: FFMPEG: fallback to use tag 0x7634706d/'mp4v'
100%|███████████████████████████████████████| 2171/2171 [04:14<00:00, 8.52it/s]
Video is ready at: dat/cars_detection.mp4
You can use the char.detect_LP or morpho.detect_LP
import cv2
from utils import plot_img
from char import detect_LP_char
img = cv2.imread("imgs/char_frame_180_car_no_lp1.png")
plot_img(img)
detected_img, LPs = detect_LP_char(img)
plot_img(detected_img)
You can further debug, and calibrate the rules via the internal functions char.detect_LP_char or char.detect_LP_morpho
import cv2
from utils import plot_img
from morpho import detect_LP_morpho
img = cv2.imread("imgs/char_frame_180_car_no_lp1.png")
plot_img(img)
detected_img, LPs = detect_LP_morpho(cv2.resize(img, (500,500)), L_min=35, L_max=60, W_min=55, W_max=90, debug=True)
plot_img(detected_img)
