TOPODRONE LIDAR 100 LITE - LIDAR POINT CLOUD DATAPROCESSING WORKFLOW

30 сентября 2020

Articles

TOPODRONE team presents its latest development - high-tech, compact, lightest and most affordable, among professional survey systems, TOPODRONE LIDAR 100 LITE.

TOPODRONE LIDAR 100 LITE is designed to perform UAV laser scanning at operating heights of about 70 - 80 meters and provides highly precision (3-5 cm x,y,z accuracy) and detailed (150 point per sq.m.) point clouds and digital elevation models.

When creating TOPODRONE LIDAR 100 LITE, we used the latest developments in the field of LIDAR & GNSS technologies and inertial measurement systems, which ensured high reliability, accuracy and ease of use.

The low weight of TOPODRONE LIDAR 100 LITE (less than 800 grams), allows the laser scanner to be installed on enterprise level drones DJI MATRICE 200/210 and DJI MATRICE 300, which provide flight times of 28 and 35 minutes, respectively, plus 20% of the battery reserve for returning home.

At the same time, it should be noted that the DJI M200 / 300 drones are among the most reliable (if not the most reliable among available commercial) unmanned systems on the market, equipped with a water proved / dust protected housing and motors, numerous sensors and obstacle avoidance systems, heated batteries , a powerful communication system that allows to be used in harsh climatic conditions, mountainous terrain, as well as in the aggressive environment of operating enterprises.

Installation of TOPODRONE LIDAR 100 LITE on the drone does not take much time, just plug the laser scanner into DJI Skyport connector or mount it on a specially designed mount on the battery compartment.

Integration with cameras DJI X4S, Sony R10C, Sony RX1RII and others is provided, which allows you to obtain highly accurate and detailed point cloud, colored in natural RGB colors.

We have specially developed several options for installing TOPODRONE LIDAR 100 LITE on the DJI MATRICE 200/210 drone, so that users can easily install our new solution on board an existing drones and integrate with the DJI X4S camera (Fig. 1 and Fig. 2 ).

Figure 1. TOPODRONE LIDAR 100 LITE installed on DJI MATRICE 200 via original DJI connector

Figure 2. Using the DJI X4S Camera and TOPODRONE LIDAR 100 LITE together on board of DJI M200

Table 1 shows main technical characteristics of TOPODRONE LIDAR 100 LITE

GNSS
Number of channels	184
GNSS systems	L1/L2 GPS, GLONASS, BeiDou, Galileo, SBAS
Data rate	Up 20 Гц
IMU
Gyro Bias Repeatability ( o /hr 1σ)	65
Gyro Bias In-run Stability ( o /hr 1σ)	3
ARW ( o /√hr )	0.15
Accel Bias Repeatability (mg 1σ)	1.0
AccelBias In-run Stability (mg 1σ)	0.02
VRW (m/s/√hr)	0.02
TOPODRONE LIDAR 100 LITE on board of DJI M200/M300
Working altitude (m)	50-70
Width of scanning corridor from an altitude of 70 meters (m)	130
Approximate point cloud density (points per sq.m.)	150
Area of survey	100
Accuracy (cm)	3-5
Weight of TOPODRONE LIDAR 100 LITE (gram)	800

Table 1.

Let’s have a look at some UAV LIDAR projects executed in cities, dense forest areas and construction sites.

In general, there are few main steps to carry out LIDAR survey project

1. Flight mission planning.

To prepare flight routs, we would suggest to use UGCS software, which has all the necessary tools for UAV LIDAR mission planning (Fig. 3). We have published several articles in our BLOG about advantages of this software for professional surveyors.

Figure 3. Mission planning for LIDAR survey considering the terrain, the configuration of the region of interest and the flight time of the drone.

2. Stage of field works.

First of all, it is necessary to set up a base station at a point with known coordinates or download Rinex files from a national survey network later

Next you just need to install TOPODRONE LIDAR 100 LITE on the drone and turn on its power. The LIDAR system is fully automatic initialized, showing its current status via LEDs.

Figure 4. Field works

Next, perform aerial survey flights in automatic mode along pre-planned routes.

At the end of the flight, insert a USB flash drive and copy the GNSS and IMU data, as well as the raw laser measurements.

3. Data processing.

There are tree main stages for LIDAR data processing

GNSS and IMU data post processing
Points cloud generating
Strip alignment, points cloud processing and classification

First of all, we would like to draw your attention to the fact that the total processing time of data from one LIDAR flight, starting from copying data from an SD card to obtaining the final point cloud, takes 30-50 minutes, depending on the skills and the performance of the computer. At the same time, all calculations can be performed on a laptop in the field, which significantly reduces the time needed to complete projects and allows you to assess the quality of dataset without leaving the area of work.

One of the first stages is GNSS and IMU data post processing. As a result you will get high-precision lidar trajectory with up to 300 Hz rate. (Fig. 5)

Figure 5. Flight Trajectory processed in Inertial Explorer

At the next step, TOPOLIDAR software is used to generate a point cloud (Fig. 6) in the user defined coordinate system, using the previously calculated trajectory of the flight and LIDAR raw dataset. (Fig. 7 - 13)

Figure 6. TOPOLIDAR software interface

Figure 7. LIDAR point cloud aligned with flight routs

Figure 8. LIDAR point cloud, colored depending on the height of objects

Figure 9. RGB point cloud

Figure 10. LIDAR point cloud, colored depending on the height of objects

Figure 11. RGB point cloud

Figure 12. LIDAR point cloud, colored depending on the height of objects

Figure 13. RGB point cloud

Point cloud classification, extracting relief surface, generating DEM and contour lines.

To visualize and classify point clouds, we recommend using LIDAR 360 software. This software has all the necessary functions for processing airborne laser scanning data, such as Strip alignment, noise filtering, point cloud editing and classification, relief surface highlighting, DEM generation. , as well as TIN models, coloring point clouds in RGB colors according to a ready-made orthomosaic, classification of vegetation and power lines, and much more.

The great advantage of LIDAR 360 software is the high speed of processing large amounts of data. For example the classification of the point cloud and the terrain extraction from one flight datset takes about 3 - 5 minutes. In Fig. 14 - 25 are examples of automatic classification of a point cloud of an area covered with dense forest vegetation.

Figure 14. Point cloud over an area covered by dense forest

Figure 15. RGB point cloud over the dense forest area

Figure 16. Terrain profile built over forested area, clearly tracing the surface of the terrain under trees

Figure 17. Contour lines merged with the point cloud, built fully automatically

Figure 18. Contour lines merged with the point cloud, built fully automatically

Figure 19. Terrain profile built over forest area, clearly tracing the surface of the terrain under trees

Figure 20. Automatically classified ground surface

Figure 21. Contours generated in automatic mode from a classified point cloud

Figure 22. Terrain profile built over forest area, clearly tracing the surface of the terrain under trees

Figure 23. Automatically classified ground surface

Figure 24. Contour lines merged with the point cloud, built fully automatically

Figure 25. Contour lines built fully automatically based on classified point cloud

The high accuracy and details of TOPODRONE LIDAR 100 LITE can significantly speed up the maps and contour lines creation for built-up and wooded areas.

In Fig. 26-31 shows an example of 3D model of a built-up area of an area of about 100 hectares, displaying buildings and structures, lighting poles and power lines, roads, fences, vegetation, obtained as a result of just one flight of a the drone with TOPODRONE LIDAR 100 LITE on board. Data processing of this survey area and generation of a highly detailed point cloud took no more than 40 minutes, which clearly shows the advantage of LIDAR technology over photogrammetric methods for constructing a three-dimensional model, and the possibility of quick and automatic classification of the relief under forest vegetation opens up new opportunities for a significant reduction in cost and the timing of field and office work, where previously only ground methods of total station survey could be used.

Figure 26. 3D model of the built-up area, showing buildings and structures, fences, power transmission towers, vegetation, roads, etc.

Figure 27. 3D model of the built-up area, showing buildings and structures, fences, power transmission towers, vegetation, roads, etc.

Figure 28. 3D model of the built-up area, showing buildings and structures, fences, power transmission towers, vegetation, roads, etc.

Figure 29. 3D model of the built-up area, showing buildings and structures, fences, power transmission towers, vegetation, roads, etc.

Figure 30. Terrain profile under forest vegetation

Figure 31 Automatically classified point cloud

In Fig. 32-35 an example of a 3D model of a construction site based on laser scanning data, which allows you to quickly monitor the amount of construction work or make value calculation after 30-40 minutes after the flight.