Community Intelligent Access Control System Of License Plate Recognition Based On LabVIEW

With the rapid development of automobile industry and the continuous improvement of people's consumption level,intelligent transportation has become an indispensable part of each road management system.License plate recognition technology in intelligent traffic system has become a hot research topic.The object of our design study is the vehicle access control system for license plate recognition community which includes software and hardware parts based on LabVIEW software platform.Because the hardware part is restricted by the experimental conditions,the emphasis of this design is only on the soft processing of license plate recognition.There are three main design works: image preprocessing,license plate positioning,character segmentation and recognition in this paper.Since there are many kinds of image processing algorithms,this design choose the best one after experimental comparison.The design method is as follows:?Image preprocessing.The image signal is transmitted to the processor(computer),when the vehicle detection device triggers the camera and through the USB bus and the image acquisition driver function of LabVIEW software.Image preprocessing is used to grayscale,enhance and filter the image which is acquired by image preprocessing.In this paper,weighted average method is used for graying.Histogram equalization is used for enhancement.Median filtering to remove noise.Finally,the image is obtained which is uniform brightness,large difference between the target(license plate area)and the background gray scaleand high definition by the above method.? License plate location.Location is to obtain license plate feature information and capture the license plate area.Firstly,the target image particles are isolated from the background particles by the iterative clustering threshold segmentation method.The license plate area is left and realized by initial positioning through this method the particles were then filtered by morphological treatment.To improve positioning accuracy,We combined the segmentation method based on the tag image and the eigenvalue method of the length-width ratio of the license plate to exclude the background particles with the same height as the target particle eigenvalue and finally accurately locate the license plate.? Character segmentation and recognition.Because of the direct segmentation and recognition of the captured license plate image will cause a large error,the image was binarized and filtered by the local threshold segmentation method and morphological processing method,and then the tilt correction was carried out by the point-coordinate correction method.Finally,the improved labview-based OCR method can be used for segmentation and recognition,and the results show that the recognition rate is improved.The results show that the image brightness is uniform and the difference between the target and the background is enhanced after pre-processing,which lays a foundation for successful positioning.The preliminary positioning and accurate positioning method can make the particle connectivity and boundary of the license plate area better,and accurately obtain the license plate characteristic value,and then the license plate area can be intercepted.Segmentation and identification are the ultimate goal.The Images which is positioned are segmented and identified,and obtains the white target character and black background through the selected algorithm.The final result shows that the characters in the image are clear without interference,connected with strokes,smooth boundary,and standard font,which makes the segmentation andrecognition accuracy over 90%.