TCS_HUMAIN

Git clone : https://github.com/AlexeyAB/darknet.git

Download the following from given links

initial weights YOLO

trained weights YOLO

training dataset

validation dataset

All other information regarding the setup can be found out in Description FILE

Different files and their roles:

1. Cropping_Plates_from_Detection_YOLOv3_for_OCR.ipynb: Crop output objects from YOLOv3 for OCR
2. Dataset1(ANPR)_preparation_for_yolo.ipynb: preprocessing ANPR dataset as per YOLOv3 training requirements
3. (Indian_cars)_preparation_for_YOLO.ipynb: preprocessing Indian Car dataset as per YOLOv3 training requirements
4. Number_Plate_Detection_YOLOv3.ipynb: Google colab notebook for training & storing final output over test data
5. OCR_for_NumberPlate.ipynb: OCR over all cropped images and storing it in a list
6. Description.pdf: Entire problem statement and steps to follow for executing the project
7. Train.txt,Valid.txt,obj.names,obj.data: Requirements for YOLOv3 as explained in description.pdf

For understanding YOLOv3 training set, click here