License Plate Recognition Algorithm Based On Set Partitioning In Bad Weather

License Plate Recognition(LPR)is the core technology and key link of Intelligent Transportation System(ITS),and is widely used in people's lives.LPR is a computer vision system based on image processing,pattern recognition and other technologies.It has the function of automatically recognizing vehicle license plates.It is simple and convenient,and has been applied in many occasions,such as car parking management system,highway toll system,and vehicle Investigation and tracking,etc.At present,there are many license plate recognition systems.However,the license plate recognition process is affected by various factors,especially the bad weather seriously affects the quality of the captured images,which increases the difficulty of license plate recognition.The effect of using general image segmentation algorithms is not Ideally,based on fuzzy C-means algorithm and kernel clustering algorithm,an adaptive segmentation algorithm based on set partitioning is proposed and applied in license plate recognition to segment the license plate image to obtain better image segmentation effect.The main work of the thesis is as follows:(1)The license plate recognition system in bad weather firstly performs grayscale and binarization on the collected license plate image.On the basis of this,combined with the color characteristics of the license plate and the edge detection method to complete the positioning of the license plate,and then set the appropriate threshold.The threshold value is used to binarize the license plate again.At the same time,the projection transformation function is used to highlight the pixel of the license plate area,and the segmentation efficiency of the license plate character is increased,and the template matching method is used to complete the recognition of the license plate.(2)According to the characteristics of fuzzy C-means algorithm and kernel clustering algorithm,a fuzzy clustering algorithm based on set partitioning is proposed to avoid the problem that the cluster center and classification number need to be manually determined.At the same time,the advantages of the nuclear clustering algorithm are combined to improve the efficiency and accuracy of the license plate recognition system.Through experiments,The FCM algorithm,AFCM algorithm and RAFCM algorithm are applied to the license plate recognition in three severe weather conditions: heavy rain,heavy snow and haze.The experimental results show that the proposed RAFCM algorithm is used to collect the license plate image in bad weather.Serious pollution,resulting in low quality of segmentation has a better treatment effect,can improve the efficiency of license plate character segmentation in vehicle identification,meet the requirements of real-time and high efficiency of license plate recognition,meet the needs of modern society for car management,and promote license plates.Intelligent development of recognition.