Skip to content

Processing Drone Imagery Using Drone2Maps

Ivy Araujo edited this page Mar 16, 2024 · 23 revisions

Aerial imagery are images obtained from airborne crafts, delivering a comprehensive aerial perspective across various scales and scopes. These images hold significant value for spatial-temporal analysis. It stands as one of the earliest methodologies for acquiring geospatial data. Over time, the traditional employment of balloons and kites has given way to more sophisticated means, such as planes, drones, and other camera systems, facilitating data collection.

There are numerous advantages in using aerial photography. Its capacity to capture details over extensive areas, coupled with its time and cost efficiency, renders it invaluable across a diverse range of industries. The resultant imagery and associated geospatial data find applications in fields such as construction, insurance, real estate, agriculture, urban planning, environmental studies, and more.

In Architecture, Engineering and Construction projects, drone imagery can be a quick snapshot in time to evaluate different conditions on site, such as change in ground conditions and the impact caused by extreme weather events. They are also a system of record to identify environment and build changes, support asset management, and assessing urban plan compliance requirements.

Basic Data Capture Requirements

When acquiring imagery data with drones, each image must be tagged with GPS coordinates, and there should be an overlap of approximately 70% or more between successive images.

image

Source: Esri MOOC - Transform AEC Projects with GIS and BIM (2024).

NADIR & Low and High Oblique Angle Imagery

• For flights aimed to produce 2D maps, the camera is vertically oriented (90 degrees) - NADIR Imagery.

• For flights intended to generate 3D products, the camera angle varies: low (0-30 degrees) or high (30-90 degrees) - Oblique Angle Imagery. Low oblique angle imagery captures a wider view of the surroundings with the horizon visible, whereas high oblique angle imagery focuses on specific objects or features, with less emphasis on the horizon. See in the table below a few application of when using each case.

Low Oblique Angle Imagery High Oblique Angle Imagery
Capturing panoramic vistas of coastal landscapes for tourism promotion. Highlighting architectural details of historic buildings for preservation documentation.
Assessing the impact of urban development projects on surrounding natural environments. Surveying archaeological sites to document specific features or excavation areas.
Documenting large-scale infrastructure projects such as highways or bridges to evaluate construction progress. Assessing damage to individual structures or landmarks after natural disasters or emergencies.
Capturing high-resolution imagery suitable for generating detailed and accurate point cloud datasets in aerial photogrammetry applications.

image

Source: Esri MOOC - Transform AEC Projects with GIS and BIM (2024).

ArcGIS Drone Solutions

Esri contains two software solutions to support drone programs: Drone2Map and Site Scan for ArcGIS:

  • Drone2Map is a desktop application primarily focuses on processing drone imagery to create 2D and 3D maps, models, and analysis. It offers capabilities such as orthomosaic creation, terrain modeling, and volumetric analysis.
  • Site Scan for ArcGIS, on the other hand, is a cloud-based end-to-end reality mapping that not only includes drone image processing but also offers mission planning, flight control, and data management features. It provides end-to-end drone mapping solutions tailored for various industries such as construction, infrastructure, and environmental monitoring.

These products are part of the ArcGIS Reality family, which is a suite of photogrammetry software products designed to enable reality capture workflows that allow the creation of accurate digital representations of the world. These digital representations can be layered with geospatial data, that enriches reality with geographical context (click here to know more).

image

Area Survey Flight Mode in Site Scan for ArcGIS: specify the region of interest and flight altitude, and Site Scan will automatically generate the flight lines necessary to map the designated area. Source: Esri MOOC - Transform AEC Projects with GIS and BIM (2024).

Processing Drone Imagery with ArcGIS Drone2Map

In this article we will briefly cover how to process drone imagery using ArcGIS Drone2Map.

  1. Open Drone2Map and start a 2D Template.
  2. Load the drone images to the project and after that click in the Create button.

image

Create a new project in ArcGIS Drone2Map and Load Drone Image.

  1. The initial view will display the locations of the images overlaid on a basemap, as each image is associated with a GPS point.
  2. In Manager, enable all 2D and 3D products to be created in the project.

image

Location point of each drone image over the basemap imagery. In Manage, enable products to be created.

  1. It is advisable to employ ground control points (GCPs) to guarantee the precision of the outputs. In this material, GCPs will be added using the imagery basemap.

For enhanced accuracy, it is recommended to utilise high-precision GPS. GCPs (x,y,z) may consist of either permanent ground features or temporary markers positioned prior to the drone flight.

  1. Turn imagery basemap on > Click on Control Manager on the ribbon > find a unique feature on the map > Click on Add Control From Map > click on the basemap to select the GCP.

It is recommended to create three or more GCPs in different areas of the covered area to generate accurate results.

GCPs are essential for refining survey accuracy, using GPS data to enhance spatial representations, especially in photogrammetry. They are crucial for precise mapping. Checkpoints, however, are optional and serve to verify survey accuracy against real-world data. Unlike GCPs, they're not integral to the survey process but provide a way to double-check accuracy by comparing model output to known coordinates collected on-site.

  1. In Control Manager, click on Show Link Image Editor. Here the selected GCP on the basemap will be linked to the equivalent CGP on the drones images in at least 3 images. The software will automatically find the tie point that connects these images to each other. However, control points are required to connect these images to the ground coordinates.

image

  1. After all GCPs are in place, click in Start to run processing. When this step is complete, the output will be generated (2D and/or 3D products).

image

image

3D point cloud output.

image

3D mesh product.

Results

  1. After the processing is complete, an output report is generated and saved in the project folder. It can be also visualised from the ribbon, at Report option (Drone2Map Training Processing Report.pdf).

image

It is recommended to analyse the processing report to look for warnings or errors. Specific items to look for in the processing report: (1) quantity of images calibrated and processed; (2) mean projection error (measured in pixels); (3) check points, which provides a measure of the accuracy of the project.

  1. Click in Link Views to simultaneously visualise 2D and 3D maps.

image

  1. After the 2D/3D outputs are generated, within Drone2Map itself, other analysis can be done, such as calculate the volume of stockpiles.

image

  1. The project can be opened in ArcGIS Pro for further analysis and processing.

image

Share Outputs to ArcGIS Online/Enterprise

  1. The outputs can be published to ArcGIS Online/Enterprise by going to the share tab. It allows these items to be shared through web maps, scenes, dashboard, experience builder and more.

image

  1. See below 👇 the YouTube overview of the Building Digital Twin displayed using ArcGIS Online Scene Viewer, including shadow and weather simulation.

IMAGE ALT TEXT HERE

Sources: