Research On Point Measurement Algorithm Of Hand-held Mobile Measuring Pole

In the construction of smart cities,the use of geographic information system(GIS)technology to build a spatial geographic database of urban infrastructure is the foundation of everything.Efficient and quick geographic information data collection is the first task to be completed.However,in a complex urban environment,signals of Global Navigation Satellite System(GNSS)are easily reflected and blocked by buildings and trees,which makes real-time kinematic(RTK)carrier phase difference technology ineffective.While traditional surveying methods such as total station surveying is time-consuming and labor-intensive.In response to this dilemma that the demand for point coordinate measurement is increasing but there is no suitable measurement method,this paper designs a hand-held mobile measuring pole based on inertial navigation technology and visual measurement technology,proposes the corresponding point measurement algorithm,and defines a set of new and efficient measurement mode,which can complete the point measurement work required by GIS in a GNSS-denied environment.The detailed research results are as follows.1.Design and build a complete set of hand-held mobile measuring pole hardware system and solution software,effectively install multi-source sensors such as Micro-Electro-Mechanical-Sensor(MEMS)inertial measurement unit(IMU),camera,GNSS antenna and trigger switch on a GNSS centering pole,and the time stamps of all sensors are synchronized to GPS time system to achieve synchronized data collection and calculation.2.Using the lever-arm compensated zero velocity update(LA-ZUPT)algorithm,an algorithm for maintaining the high-precision pose of the measuring pole in a GNSS-denied environment is proposed.By holding the pole and walking in crutches mode,the tip of the pole regularly touches the ground to obtain the shortterm zero velocity correction condition,which effectively suppresses the divergence of errors of MEMS-IMU.At the same time,based on the principle of trajectory similarity matching,an algorithm that uses the first and last control points to initialize the yaw in a GNSS-denied environment is implemented.3.Design the measurement method of only using the inertial measuring pole to measure the ground points to be measured in a GNSS-denied environment,and verify the effectiveness of the method through multiple sets of experiments under different conditions.The experimental results show that when the trajectory geometric conditions are good,that is,the control points are distributed on both sides of the point to be measured,and the distance between the measured point and the control point is less than 20 m,the plane point accuracy can reach 0.1m.4.In a GNSS-denied environment,the pose provided by the inertial measuring pole is used in conjunction with vision technology to measure the coordinates of the point to be measured,and 4 data processing modes are proposed and designed:direct triangulation;joint optimization of the camera pose and the point to be measured;optimization of adding control point constraints;applying image geometric constraints to refine the pose and then triangulating.The actual measurement results show that whether absolute point measurement accuracy or relative measurement accuracy,the best result is optimization result of adding the control point constraint.Almost all the point position errors are within 0.1m,and the RMS of the length errors of the measured line is only 2cm.As for the direct triangulation and joint optimization of the camera pose and the point to be measured,the difference of the measurement effect between two processing methods is very small.The RMS of absolute point position errors is less than 0.2m,and the RMS of line length errors is less than 4cm.Using geometric constraints between images to refine the camera pose and then triangulating,the absolute point position accuracy is improved,and the absolute point position errors of most(85.4%)measured points are less than 0.2m.This paper designs and implements an inertial/visual mobile measuring pole,proposes the corresponding point measurement algorithm,and defines two measurement modes of the pole tip touching the ground point to be measured and the visual non-contact measurement.In a GNSS-denied environment,the point measurement accuracy can basically reach within 0.2m,which can meet the requirements of coordinate data collection of most GIS applications.