Automatic Number Plate Recognition using Region-based Convolution Neural Network and LSTM
Automatic Number plate recognition is the most extensive practice used in diversified ways to identify an image of vehicles number. This practice has undergone many technologies to overcome significant drawbacks over the years. One of the major concerns observed is the recognition and sequencing of the number plate. Although the identification is not a primary apprehension, the object recognition and order of the number plate is the most important alarm.
So considering this problem the present paper deals with R-CNN and LSTM to identify and recognition of an image of the number plate. In this method, the arbitrary sequence of the number and the exact text analysis in a particular format is safeguarded. The combination of both Convnet and Recurring Neural Network with LSTM result in a accurate numbers of vehicles.