Introduction

Mobile LIDAR technology is increasingly used for large-scale projects due to its significant productivity levels compared to traditional capture methods.

The basis of mass captures in motion is the implementation of inertial sensors, which, by measuring the accelerations and angular velocities of the system, allow for obtaining a relative position of the mobile device. However, to determine absolute trajectory positions, the incorporation of GNSS receivers is necessary. This combination creates a redundant system that provides a robust solution and extremely high precision in this type of work.

Mobile LIDAR technology provides highly representative solutions with a high level of precision. This total precision is defined by the relative precision of all its components. A very relevant aspect is the environment in which the capture is performed, as it is a factor that affects the precision provided by GNSS.

In this document, we will explore the use of ground control points, which will improve the overall accuracy of the captured data.

 

  

Accuracy of Mobile LIDAR Systems

The absolute accuracy of a point cloud captured using a mobile LIDAR system is defined by the relative accuracy obtained by the inertial system plus the GNSS receiver, a combination known as INS. The accuracy of the LIDAR + INS capture sensor results in the absolute accuracy of the survey.

One of the key factors related to the absolute accuracy of the captured point cloud is the GNSS trajectory information, as it will provide the correct spatial location of the LIDAR information. This location is mainly affected by environmental factors in the study area, regarding satellite coverage and obstructions that affect sky visibility above the mobile device.

To improve the GNSS solution in complex environments, as described above, an accessory called DMI can be incorporated into the mobile LIDAR system. This accessory is mounted on the vehicle's wheel and basically functions as an odometer that will assist the INS when satellite coverage is lost.

Additionally, to improve the final result of the LIDAR information, it is possible to incorporate control points captured in the field. From these, the trajectory is adjusted to bring it to the reference location.

Control Points

Control points are marks located on the ground that are identifiable in the point cloud and will be measured using traditional surveying instruments to assign coordinates in the project's reference system.

The location of these marks is directly related to the project's objective and capture scenario, as well as the precisions associated with the inertial sensor. The higher the precision of the inertial sensor, the further apart these points can be spaced, up to approximately 500m, located on both sides of the corridor.

The geometry of the card defining the control point location will depend on the user. Marks can be made in a defined shape (squares, rectangles) with reflective paint on the pavement, painting black/white areas, defining the point with coordinates in the center of it, or simply assigning coordinates to existing pavement marks in the area that meet the necessary distribution.

Fig 1. Control points 

 

Regarding the capture accuracy of these points' coordinates, it depends on their objective, ranging from merely having self-control of the survey to higher levels of accuracy where geometric leveling will be necessary to perform the altimetric corrections of the project.

 

LIDAR Data Registration

Once the trajectory calculation processes for the survey, its validation, and point cloud generation have been completed, it is possible to register the LIDAR information against the coordinates of the control points captured in the field.

To perform this process, the software applications have their own adjustment interfaces. In this case, we will review how Trimble performs this process using Trimble Business Center software with its Mobile Mapping module.

In this registration module, the coordinate of the control point is associated with the mark in the point cloud. For this, the processing software uses algorithms that adjust the trajectory, making corrections to align the point cloud with the captured topographic information.

 

Fig 2. Control Point Registration

 

This algorithm offers alternatives to be more or less restrictive in adjusting the resulting cloud; each is recommended for specific applications. The most common option is to perform a "Global and then local" registration, thereby adjusting the trajectory to the control points while avoiding deformations at the route boundaries.

 

Fig 3. Trajectory Registration Methods

 

Registration Results

In this project, the scenario did not present significant challenges regarding the satellite environment, and the initial accuracies of the mobile LIDAR survey already met a good standard for many applications. However, among the applications for this technology, some may require an even higher level of precision.

The control points were distributed on each side of the route, interspersed approximately every 250m, and measured with high precision. During registration, alternating data was used. For the purpose of this exercise, self-control points were used; however, in reality, all topographic point data could have been used to calculate the trajectory adjustment.

Once the adjustment is made, the following adjusted values and their residuals are observed, according to the assigned control point.

 

Fig 4. Control Point Residuals

 

Finally, to validate the adjustment process, an analysis of the self-control points is performed, yielding the following results:

 

Fig 5. Analysis of self-control points 

 

Conclusions

One of the most important factors when capturing information with Mobile LIDAR systems is having sufficient satellite availability to obtain good accuracies. In more complex environments, the contribution of the DMI sensor is very significant when performing trajectory calculations.

The accuracies obtained directly by calculating the trajectory of the Mobile LIDAR system allow for good results for most applications. However, in other applications that require higher levels of accuracy, the final results can be significantly improved by registering the LIDAR information against a network of control points measured with a higher level of precision.

The workflow presented by Trimble Business Center in its Mobile Mapping module helps the user manage large amounts of LIDAR data and improve the pressures of the surveys performed, ultimately obtaining final results for vectorization and calculation of other parameters that we will review in other publications.