How does the mapping of remote territory using drones
The Swiss company SenseFly demonstrated a successful method of mapping inaccessible areas with the help of drones, showing that in a limited time it is possible to obtain a highly detailed model of a complex relief of a large area that can be published almost immediately on the Internet.
For mapping was chosen peak Matterhorn height of 4478 meters, located in the Swiss Alps on the border between Italy and Switzerland (this place on Google Maps). Technically, the process itself looked something like this: a team of engineers, being at the top of the peak, launches several eBee drones, controlling their flight along a given route using a tablet using GPS. The aircrafts of this class themselves can perform aerial photography of an area of up to 10 km 2 (from one flight), allowing at the post-processing stage to receive maps and terrain models with an accuracy of 5 cm - despite the fact that the camera installed on board the drone is modest by the standards of some modern phones in 16 Mp. The wingspan of the drone is a little less than one meter - 96 cm, and the weight is 700 grams.
As a result, in order to obtain the model of the peak Matterhorn, drones needed to perform 11 flights, reaching a maximum height of 4,707 meters and flying a total of about 264 km. All the time aerial photography took 5 hours and 40 minutes, during which 2188 images were taken.
How this process looked from the side of the team of surveyors, you can see the video below:
')
The post-processing stage of the results consisted in the visualization of the obtained data. To do this, from the internal LAZ format (compressed format for describing point clouds), the data was converted to standard ASCII (using LASTools tools), and then visualized using the XP-Points JavaScript library, which can render points as WebGL models.
The result can be viewed on the gif-image under the spoiler, or go directly to the interactive model (requires a browser with WebGL support) - it is noteworthy that 300 million dots are displayed (the size of the source file with them is 26682.732 Kb):
Also on the mosaic image of the shooting results, the difference in the image quality between the existing aerial photography and the obtained robotized methods is clearly seen:
It is noteworthy that all GIS tools used for processing are free and freely available on github .
[ Source ]
For mapping was chosen peak Matterhorn height of 4478 meters, located in the Swiss Alps on the border between Italy and Switzerland (this place on Google Maps). Technically, the process itself looked something like this: a team of engineers, being at the top of the peak, launches several eBee drones, controlling their flight along a given route using a tablet using GPS. The aircrafts of this class themselves can perform aerial photography of an area of up to 10 km 2 (from one flight), allowing at the post-processing stage to receive maps and terrain models with an accuracy of 5 cm - despite the fact that the camera installed on board the drone is modest by the standards of some modern phones in 16 Mp. The wingspan of the drone is a little less than one meter - 96 cm, and the weight is 700 grams.As a result, in order to obtain the model of the peak Matterhorn, drones needed to perform 11 flights, reaching a maximum height of 4,707 meters and flying a total of about 264 km. All the time aerial photography took 5 hours and 40 minutes, during which 2188 images were taken.
How this process looked from the side of the team of surveyors, you can see the video below:
')
The post-processing stage of the results consisted in the visualization of the obtained data. To do this, from the internal LAZ format (compressed format for describing point clouds), the data was converted to standard ASCII (using LASTools tools), and then visualized using the XP-Points JavaScript library, which can render points as WebGL models.
The result can be viewed on the gif-image under the spoiler, or go directly to the interactive model (requires a browser with WebGL support) - it is noteworthy that 300 million dots are displayed (the size of the source file with them is 26682.732 Kb):
Matterhorn
Also on the mosaic image of the shooting results, the difference in the image quality between the existing aerial photography and the obtained robotized methods is clearly seen:
It is noteworthy that all GIS tools used for processing are free and freely available on github .
[ Source ]
Source: https://habr.com/ru/post/199194/
All Articles