Vehicle License Plate Recognition Based on Edge Detection

A vehicle license plate recognition system is proposed. The focus is on a new method for detecting number plates. The method is based on the detection of contours and the use of geometric information about the contours. To detect the contours, a mathematical image model based on a two-dimensional discrete-valued Markov chain with two states was used. To reduce computational resources, we propose to carry out detection of number plates by the bit planes of the senior, most informative digits of the digital image. The proposed method allows to significantly reduce the computational cost of detecting the license plate and, accordingly, on the recognition of vehicle license plate symbols. Character recognition accuracy reaches 92%.