Research On License Plate Recognition Under Complex Conditions Based On Machine Vision

With the rapid development of China's economy,traffic congestion has become a common phenomenon in many cities in China,especially in holiday tourism,traffic congestion and frequent traffic accidents.With the development of urban digitalization and the proposal of driverless technology,intelligent transportation system emerges as the times require.With the digital development of cities and the introduction of driverless technology,intelligent transportation system emerges at the right moment.Replacing human vision with machine vision can not only improve work efficiency and reduce labor intensity,but also save a lot of manpower,material resources and financial resources,so as to make urban transportation develop more intelligently.This paper studies the license plate character recognition under the complex conditions of high-speed driving and dark light at night.Through Wiener filter restoration and image enhancement technology,the low-quality image can be effectively transformed into the image that meets the requirements of subsequent experiments.License plate recognition mainly consists of license plate image acquisition,license plate location,character segmentation and character recognition.In this paper,the industrial camera is used to obtain the license plate image,and the MATLAB software is used to preprocess the image.The method of pixel statistics is used to locate the license plate.The algorithm of vertical projection and template matching is used to segment the characters.Finally,an algorithm combining neural network and template matching is proposed to complete the license plate character recognition.Experiments show that the recognition algorithm in this paper can effectively improve the accuracy of character recognition,and can complete recognition under complex conditions.