Development of Standard License Plate Recognition for Vehicles

In the field of vehicle identification, number plate recognition technology holds a key role in intelligent building security systems. This design employs a deep learning method to achieve precise number plate recognition. The process includes several processes, including character segmentation, identification, and number plate area localization. The design takes the letters from the processed number plate image and turns them into text using Optical Character Recognition (OCR). The extracted text is then conveniently stored in an Excel document for future purposes and reference. The datasets for the Convolution Neural Network (CNN) training were obtained by manually taking the image of plate number itself. For efficient training with high accuracy, the plate number has 225 image images taken for training. The website application is used to operate the classification process using localhost. The result of the classification process shows that it has high accuracy of 90%. This study strengthens the idea of advancement in artificial intelligence innovation which could bring more usage for the betterment of the land transport sector.