Research On Intelligent Recognition System Of Parking Space Based On Laser Radar

Parking space recognition is the front key technology to realize automatic parking,in today’s automatic parking system,on the test success rate is not very desirable,still need to manually choose parking type,did not achieve very high intelligence,is not enough to adapt to the changeable environment,such as car parking problem,now in this paper,a high degree of intelligent parking recognition system,to improve the car to identify the environmental adaptability of problem has important reference value.After analyzing the advantages and disadvantages of various sensors,this paper selects lidar as the environment sensing sensor to design a clustering method of vehicle contour data suitable for parking scene,which improves the existing feature segment extraction method and increases the adaptability in parking spot recognition scene.The parking space model is built,and the relevant parameters of the target parking space are extracted from it,and the fuzzy control system is established.The key parameters of the target parking space are taken as input to establish fuzzy rules,and the parking space identification results are output at last.The main contents are as follows:Aiming at the problem of error detection and poor stability of traditional filtering algorithm when processing the original data of parking Spaces,a set of filtering processing algorithm suitable for radar ranging data and improving the reliability was designed.The original data is classified according to the principle of majority by minority,and then the difference between adjacent data is calculated,and the maximum difference value is taken as the data cut-off point.In view of the traditional data clustering method in reducing vehicle contour feature of not accuracy,in this paper,combined with laser radar scanning principle,the outline of the relative position of vehicle,vehicle car characteristics and the characteristics of laser radar data,considering the existing adaptive threshold clustering algorithm to adapt to the requirements of this article,then carries on the optimization improvement,based on the initial density,and joined the clustering process.A comparative experiment is designed to prove that this method can effectively restore the parking space contour.Line extraction algorithm based on the traditional on Line extraction,frequent Line through mergers and over-segmentation phenomena,this article will Support Vector Machine(Support Vector Machine,SVM)algorithm is introduced into the traditional Iterative endpoint Fitting(Iterative End Point Fitting,IEPF)algorithm and Line tracing(Line Track,LT)algorithm of automatic tagging classification,to optimize the selection of threshold value method,effectively improve the segmentation and merging.A real vehicle experiment is designed to verify the algorithm,and the experiment proves that the algorithm achieves the desired effect.A fuzzy reasoning system is established to realize the intelligent recognition of parking Spaces.Firstly,the parking space model was established,and several key parameters were calculated and extracted as the input of the fuzzy reasoning system.The fuzzy rule base was established according to a variety of parking scenarios and expert experience,and the parking space type was taken as the output.Prescan simulation software is used to establish a variety of parking scenarios,and experiments based on the proposed algorithm and the traditional algorithm are carried out respectively.The experiments prove that the proposed algorithm has good environmental adaptability and effectiveness,and can make the parking space recognition more intelligent.In conclusion,the intelligent identification system of parking space based on laser radar is designed.The application results show that the system can accurately identify the traditional parking space and irregular parking space,and has a high degree of intelligence and good adaptability in various scenarios.