Number Plate Recognition Based on Multi Feature and Character Recognition

Abstract

Intelligent transportation system has become an integral part of the transportation system and it consists of number plate recognition. A recognition system implemented to control the payment of fees in parking areas, highways, airport or harbour cargo control etc. There are considered an approach to identify the vehicle through recognizing of its license plate using multi feature recognition and neural network techniques. Multi feature recognition is to enhance several features of for target. The proposed method use improved character recognition by through setting proper numbers of hidden layers in the character kinds. This method can improve the efficiency and accuracy of vehicle number plate recognition. Multi pattern recognition is an algorithm for recognition of multi feature concept into the information. Proposed method solves the number plate with colour features as well as texture features and back ground features. By using multi feature concept problem of the ambiguity of character recognition can be solved. After using multi feature concept for number plate positioning, it can be obtained the projection of histogram mainly through a threshold. The result of number plate recognition is to change the plate character in to text character. Using back propagation of neural networks to resolve complex characters number plate and defective characters, and avoid heavy data analysis and mathematical modelling works. Recognition problems of automobile number plate such as character kinds, the numbers, and confusing of words can also be solved Keywords: fuzzy theory; number plate recognition; character recognition.