License Plate Recognition using a Sequential Model and OpenCV

With the increasing number of vehicles on the road every day, tracking every vehicle can be difficult in today's world. Due to the tremendous diurnal development of vehicle assiduity, it is not possible to manually keep a record of the entire vehicle density on the road. Many techniques like MATLAB, OCR (Optical Character Recognition), and YOLO (You Only Look Once) were used. This study proposes an automatic vehicle detection system that makes use of a surveillance camera. With the advancement of this system, it is now simple to keep a record and use it when needed. To identify the characters, the CNN model is trained and used. The extracted images are then segmented. This system is primarily intended for security purposes as safety and security in corporate places is a challenging task. As soon as the car arrives at the security checking area, the system will snap an image of it. The segmentation procedure is then used to extract the collected images. The data obtained is then compared to the data in their database. Here, a model has been constructed for training by using a sequential model from TensorFlow, Keras, and models, and the model produces the best results at epoch 23.