Image Processing Applications In Vehicle Type Recognition

The intelligent traffic system, image processing was studied based on the theoryof vehicle type recognition technology. The original image preprocessing and vehicletype recognition technology is one of the major part of. Based on thereview of theliterature, the following several aspects to make a detaileddescription and research:Image preprocessing: including gray image processing, binarization, noiseremoval, edge detection.. The quality of image pretreatment can improve the image,acquire more accurately for corner vehicle type recognition information.Extraction of feature extraction models: model characteristics will directly affectthe recognition results. Different characteristics of different models in the value is todistinguish the main basis for the vehicle type, this paper introduced the theory ofvisual invariants in projective invariants and cross ratio concept, the vehicleroofmidpoint and body four endpoints of the five corner points as featurepoints,computing the projective invariants to identify the type of vehicle.Research of vehicle type recognition method naive Bayes classification model:the experiment will be affected by the noise uncertainty, causing the vehicletypejudgment. This paper introduces naive Bayes classification model, in thecornerinformation is inaccurate or even wrong conditions, determine the probabilityof vehicles belonging to category of value, by comparison, selection probabilityvalueof the largest category as the type of vehicle, improve the accuracy of vehicle typeclassification.This paper has made some progress in the research work of vehicletyperecognition, but on how to reduce the effect of noise on recognition, how toaccurately obtain the corner point information and other aspects also need furtherimprovement, if you can do this for different types of vehicles the sametypes ofidentification will be more practical significance.