Research On Target Automatic Measurement Method Based On Internet Street View Data

With the rapid development of spatial information technology and computer internet technology,street view maps are increasingly used in daily life.Street view services provided by Google and Baidu have covered seven continents.Compared with points,polylines,and polygons in traditional maps,panoramic images of street view maps provide users with more intuitive and abundant information.However,these massive and free internet street view data still remain in the browsing and display stage.How to mine and make full use of these data has become a research hotspot.The most important thing is the scalability of the target in street view data.However,due to the sparsity of the laser point cloud in the internet street view data and the lack of strict internal and external orientation elements in the panoramic image,it is difficult to achieve reliable measurement of the target in the internet street view data.In order to solve these problems,this paper proposes an automatic measurement method based on internet street view data.This method can effectively use the limited information provided by internet street view data to achieve a reliable measurement of the target.The research content and major achievements of this article include:(1)Two-view-based panoramic target measurement methodThe method selects two different panoramic images,from which the target pixel position in the panoramic image,the orientation of the street view collecting vehicle when shooting the panoramic image and the geographical location of the camera,construct a straight line where the viewpoint,image point and object point are located;According to the orientation of the target,an appropriate projection surface is selected,and the forward intersection is used to solve the target geographical position.Experimental results show that this method can achieve reliable measurement of different orientation targets in street view data.(2)Single-view-based panoramic target measurement methodThe method is mainly based on a single panoramic image extraction depth of each target,combined with the relative orientation of the target in the geographic space to achieve the measurement of the target.The experimental results show that this method can effectively solve the problem that the target that can only be seen in a single panoramic image cannot be measured based on two viewpoints.The accuracy of the measurement result is slightly lower than that of the two viewpoints,but it can still meet the daily needs of application requirements.(3)Automatic Measurement Method for Panoramic Targets under Various ConstraintsThis method is based on the two-view-based panoramic target measurement method.It introduces multi-view constrained and repeated measurement targets,and introduces target geometric information constraints and depth value constraints to eliminate a large number of pseudo-intersections existing in multi-view and multi-target forward intersections.The single-view panoramic target measurement method performs supplementary measurements on targets that can only be seen in a single panoramic image.Experimental results show that the proposed method can effectively solve the problem of pseudo-intersection and target repetitive measurement caused by multiviewpoint,multi-target forward intersections in the process of target automatic measurement,and at the same time avoid the leakage measurement of the target.The research results show that the proposed method based on internet street view data can achieve reliable and rapid measurement of internet street view data and further expand the effective use of existing internet street view data.