Research And Application Of Location Recognition Method Based On Digital Image Technology

In recent years,as China has entered a new era of socialism,the scale of cities has increased,and the number of family cars in the city has grown exponentially.Therefore,the problem of urban road traffic and urban public parking spaces has become increasingly prominent;so the issue of urban traffic has become a hot topic of research.One of the research directions in this field is the automatic identification of parking spaces.Based on the concept of “Internet accelerated speed” and the concept of smart city,this paper adopts technologies such as the Internet of Things and digital image processing,takes urban parking spaces as research objects,and designs and implements a wireless parking scheme of status recognition based on digital image processing.There are three main problems to be solved in this paper: rapid acquisition and wireless transmission of multi-vehicle real-time image,splicing and detection of parking spaces,and determination of parking space usage.First of all,in order to be able to monitor more information about parking space,multiple points of continuous shooting is required,and multiple adjacent photos are collected as the original image for image stitching.Then,using the short-range wireless transmission technology,transmitting the original pictures to the PC for further processing.Then,pre-processing the original image,such as image splicing,image filtering,image enhancement,corrosion expansion,finding the max-connected area has used.The purpose of this series of processes is to detect parking space,because in the stitched image,this parking area consists of the max-connected area.The parking space can be identified by training the SVM classifier at the corner of the car.Finally,the use of the license plate in the parking space is detected to determine the use of the parking space.Among them,the license plate's monitoring and positioning is also using SVM support vector machine method to train the classifier.Each process in the experiment is basically presented in the form of data and experimental effect diagrams.The experimental results show that the proposed method is feasible and effective.This article mainly uses the Raspberry Pi hardware platform for embedded development,and uses the PC to do image processing related development.The Internet of Things wireless transmission technology and digital image processing technology is combined,and the key technologies in the experimental process are compared and studied.The research results have positive significance for the development of Internet of things and the problem of digital image detection for parking spaces.