Project that uses Yolov8 as license plate detector, followed by a filter that is got selecting from a filters collection with a code assigned to each filter and predicting what filter with a CNN process
Download all the files of the project to a folder, unzip the zip files.
All the modules necessary for its execution can be installed, if the programs give a module not found error, by means of a simple pip.
The most important:
paddleocr must be installed (https://pypi.org/project/paddleocr/)
pip install paddleocr
yolo must be installed, if not, follow the instructions indicated in: https://learnopencv.com/ultralytics-yolov8/#How-to-Use-YOLOv8?
pip install ultralytics
Are attached, te best.pt file, that allows the license plate detect for yolo( may be get from scratch following the steps pointed out in https://github.com/ablanco1950/LicensePlate_Yolov8_Filters_PaddleOCR), and FSRCNN that allows the working of the filter with de same name
As a previous step, the X_train and the Y_train that the CNN needs are created, the X_Train is the matrix of each image and the Y_train is made based on the code assigned (from 0 to 10) to the first filter with which paddleocr manages to recognize that license plate of car in the reference project https://github.com/ablanco1950/LicensePlate_Yolov8_Filters_PaddleOCR.
The Crea_Xtrain_Ytrain.py program is attached (its execution is not necessary), whose result after applying it to different image files (the input file is indicated in instruction 15) of renamed image cars with their registration plate is saved in the Training folder , consisting of the image itself and a .txt file with the name of the car's license plate and containing the filter code assigned to that image. This Training file that is attached in .zip format, is necessary to download it, like the Test.zip file and unzip.
You may run
TrainCodFilterCNN.py
but is not necessary because the files with the weights and the necessary and verified information are attached: FilterCNN16Hits500epoch.json and FilterCNN16Hits500epoch.h5. Since the CNN training process can give different results in each execution of the training.
So you can run directly
GetNumberInternationalLicensePlate_Yolov8_CNNFilters_PaddleOCR_V1.py
Any folder may be tested changing instruction 14, the resultas in LicenseResults.txt file
Comparing with the reference project: https://github.com/ablanco1950/LicensePlate_Yolov8_Filters_PaddleOCR you get a lower precision but a considerable reduction in execution time.
The references are identical to those of the project https://github.com/ablanco1950/LicensePlate_Yolov8_Filters_PaddleOCR:
References:
https://pypi.org/project/paddleocr/
https://learnopencv.com/ultralytics-yolov8/#How-to-Use-YOLOv8?
https://public.roboflow.com/object-detection/license-plates-us-eu/3
https://docs.ultralytics.com/python/
https://medium.com/@alimustoofaa/how-to-load-model-yolov8-onnx-cv2-dnn-3e176cde16e6
https://machinelearningprojects.net/number-plate-detection-using-yolov7/
https://github.com/ablanco1950/LicensePlate_Yolov8_MaxFilters
https://github.com/mrzaizai2k/VIETNAMESE_LICENSE_PLATE
https: //www.doubango.org/webapps/alpr/
Filters:
https://github.com/Saafke/FSRCNN_Tensorflow/tree/master/models ( downloaded module FSRCNN_x4.pb)
https://learnopencv.com/super-resolution-in-opencv/#sec5
https://learnopencv.com/super-resolution-in-opencv/
https://gist.github.com/endolith/334196bac1cac45a4893#
https://stackoverflow.com/questions/46084476/radon-transformation-in-python
https://gist.github.com/endolith/255291#file-parabolic-py
https://learnopencv.com/otsu-thresholding-with-opencv/
https://towardsdatascience.com/image-enhancement-techniques-using-opencv-and-python-9191d5c30d45
https://blog.katastros.com/a?ID=01800-4bf623a1-3917-4d54-9b6a-775331ebaf05
https://programmerclick.com/article/89421544914/
https://anishgupta1005.medium.com/building-an-optical-character-recognizer-in-python-bbd09edfe438
https://datasmarts.net/es/como-usar-el-detector-de-puntos-clave-mser-en-opencv/
https://felipemeganha.medium.com/detecting-handwriting-regions-with-opencv-and-python-ff0b1050aa4e
https://github.com/victorgzv/Lighting-correction-with-OpenCV
https://medium.com/@yyuanli19/using-mnist-to-visualize-basic-conv-filtering-95d24679643e