Reality Mapping vs Digital Twins

This Esri article sounds like a lot of marketing word salad, but the animated GIF is cool!

True Orthos vs Trad Orthos & ArcGIS Reality

First, what is ArcGIS Reality?

I'm not entirely sure, here's what Esri says: ArcGIS Reality is a suite of photogrammetry software products designed to enable reality capture workflows for sites, cities, and countries.

One main feature seems to be the abilaity to create & use so-capped True Orthophotos as opposed to Traditional Orthophotos.

• True Orthophotos vs. Traditional Orthophotos: Why do pixels matter in the age of AI?
• The Different Types of Ortho Images Created in ArcGIS

The animated GIF below demonsrates the main difference between true orthos & trad orthos. 

iPhone LiDAR

Hmm, maybe the excuse I need to upgrade my iPhone: Evaluation of the Apple iPhone 12 Pro LiDAR for an Application in Geosciences.

Drones vs Laser!

Eker et al. (2019) Monitoring of Snow Cover Ablation Using Very High Spatial Resolution Remote Sensing Datasets is a great paper comparing super-detailed snow cover mapping using drones and terrestrial laser scanning and includes detailed methods and processing information.

Snow Depth Mapping using SfM and more

Goetz and Brenning (2019): Quantifying uncertainties in snow depth mapping from structure from motion photogrammetry in an alpine area. Water Resources Research, doi: 10.1029/2019WR025251.

Filhol et al. (2019): Time-lapse Photogrammetry of Distributed Snowdepth During Snowmelt. Water Resources Research, doi: 10.1029/2018WR024530. Ground-based, oblique with 3 cameras - easy in concept, but the details are complex. They make a good point: usually we either have high-frequency time series from point measurements or sporadic spatial data from satellites, UAVs, etc. Their approach tries to bridge that gap.

LiDAR vs Photogrammetry

Drone LiDAR or Photogrammetry? Everything you need to know. is an interesting and useful read. But it misses the point (I think). LiDAR is LIDAR = gives you a DEM, DTM, DSM - whatever you prefer to call it. Photogrammetry gives you orthophotos - which is something entirely different. Unless you are talking about doing SfM from your images.

ArduPilot Autopilot Suite

The ArduPilot Autopilot Suite is everything and more you need for any kind of autonomous vehicle. More also here.

Drone Resources

A simple compilation of papers and other resources related to drones and their applications.

• Rossini et al. (2018): Rapid melting dynamics of an alpine glacier with repeated UAV photogrammetry. Geomorphology, 304, 159-172. UAVs and SfM.
• 8 tips that will let you achieve survey-grade accuracy of your drone survey.
• Wigmore and Mark (2017): Monitoring tropical debris-covered glacier dynamics from high-resolution unmanned aerial vehicle photogrammetry, Cordillera Blanca, Peru. The Cryosphere, 11, 2463-2480.
• Ryan et al. (2017): Derivation of High Spatial Resolution Albedo from UAV Digital Imagery: Application over the Greenland Ice Sheet. Frontiers in Earth Science, May 2017. Using a basic consumer grade camera to measure albedo. Not quite as simple as it sounds, but very cool!
• Dabski et al. (2017): UAV-based detection and spatial analysis of periglacial landforms on Demay Point (King George Island, South Shetland Islands, Antarctica. Geomorphology, 290, 29-38. Home-made professional fixed-wing drone paired with a Canon 700D DSLR.
• O'Connor et al. (2017): Cameras and settings for aerial surveys in the geosciences: Optimising image data. Progress in Physical Geography, 41(3), 325-344. Lot's of details about cameras and lenses.
• Krsak et al. (2016): Use of low-cost UAV photogrammetry to analyze the accuracy of a digital elevation model in a case study. Measurement, 91, 276-287. Using a DJI Phantom 2 Vision+ to create a high-quality DEM.

Comparing SfM and LiDAR

Nice paper by Cawood et al. (2017) comparing the accuracy of outcrop mapping using traditional methods, terrestrial LiDAR, terrestrial SfM, and airborne SfM.

• Terrestrial SfM: 446 images from 20 camera positions, issues with occlusion = parts of the outcrop are not visible
• Airborne SfM: 202 images, 6.24 mm resolution, no issues with occlusion.

Overall, SfM worked better than LiDAR (and of course does not require any specialized hardware). Issues arise from occlusion, different lighting, and low-contrast surfaces.

Drones at High-Altitude

Drones are being used extensively for scientific research these days, but there are only a few examples from high-elevation areas (e.g. 4,000 m and higher). The main challenge seems to be dilemma of battery power, flight time, and the thin air. This seems to more of an issues for multi-rotor drones as opposed to fixed-wing drones. The thinner air means that the UAV has to fly faster and is therefore less stable. In addition it can be difficult for a fixed-wing UAV to get enough lift for take-off - thus a stronger engine and/or larger wings would be helpful.

The basic workflow is simple: the UAV takes overlapping images that include their location based on internal GPS. Those are then converted in combination with GCPs into high-resolution DEMs (via SFM) and ortho-mosaics.

• Monitoring Tropical Debris Covered Glacier Dynamics from High Resolution Unmanned Aerial Vehicle Photogrammetry, Cordillera Blanca, Peru (Wigmore and Mark, 2017).
• Seasonal surface velocities of a Himalayan glacier derived by automated correlation of unmanned aerial vehicle imagery (Kraaijenbrink et al. 2016). Some of their flights reached 5,900 m.
• USING UNMANNED AERIAL VEHICLES FOR GLACIER MONITORING IN THE HIMALAYAS (Immerzeel et al. 2016).
• High-resolution monitoring of Himalayan glacier dynamics using unmanned aerial vehicles (Immerzeel et al. 2014).

UAV Mapping of Greenland Glacier

Here's another great example of a UAV used in glaciological research - all for under $2,000 and the authors provide a detailed description of their methods.

• Ryan et al. (2015): UAV photogrammetry and structure from motion to assess calving dynamics at Store Glacier, a large outlet draining the Greenland Ice Sheet.
• Skywalker X8 UAV airframe
• Ardupilot APM autopilot (open-source)
• Agisoft Photoscan Pro (for the 3D generation)

Jouvet et al. (2017) provided a more recent example using the same system and workflow to generate 10 cm orthoimages of the Bowdoin Glacier in Greenland.

Ryan et al. (2014, their Figure 2)

Drones map snow depth

This is new (to me): Harder et al. (2016) used a senseFly eBee RTK UAV to map snow depth. In essence you fly first over snow-free terrain and create a DSM, then create a second DSM over snow cover and subtract the two - clever. There are some limitations (e.g. snow depth has to be greater than 30 cm), but otherwise this seems quite feasible.

LiDAR and SFM

Everything you always wanted to know about LiDAR and SFM: Analyzing high resolution topography for advancing the understanding of mass and energy transfer through landscapes: A review (by Passalacqua et al., 2015, Earth-Science Reviews).

The title is a bit misleading...the focus is very much on airborne LiDAR. Plus, the term 'transfer' implies a process (or processes), but LiDAR and SFM can only give you 'states' at high spatial resolution, but (in-practice) not high enough temporal resolution to get at actual processes, transfers, and rates thereof.

Passalacqua et al. (2015, Figure 2)

Drones and SFM inside a volcano!

Worth the annoying Geico ad...

UAV Mapping of Himalayan Glaciers

Here's another UAV glacier-mapping example: Immerzeel et al. (2014). As they correctly point-out UAV imagery can be the missing link between point measurements and space-based sensors. The video below is an overview of their research (look for the UAV at about 4:00 minutes).

• Swinglet CAM UAV
• Canon IXUS 125 HS camera
• Structure-from-Motion using Agisoft Photoscan Professional
• Case study by senseFly

UAV and Structure-from Motion

There is more high-tech in this paper than I can handle: Lucieer et al. (2014) Mapping landslide displacements using Structure from Motion (SfM) and image correlation of multi-temporal UAV photography. Progress in Physical Geography 38(1), 97-116.

• Here's the Mikrocopter UAV they used.
• Structure-from-Motion was implemented using Agisoft (see also this previous post about Structure-from-Motion).
• The displacement was determined using COSI-Corr from CalTech.

UAVs in Science and Glacier Research

Here is an actual practical example of using a UAV in a real scientific application - not just to get some cool pictures and video, but to construct a high-resolution DEM for a remote glacier in the Canadian Arctic.

• Whitehead et al. (2013, The Cryosphere)
• Additional Information
• Outlander/CropCam UAV
• Inpho Software

Here's a nice review paper: UAVs as remote sensing platform in glaciology: Present applications and future prospects (Bhardwaj et al. 2016, Remote Sensing of Environment 175).

Structure from Motion

Structure from Motion (SfM) refers to algorithms by which you can calculate - based on many overlapping images - both the relative camera geometry (motion) and the 3D structure captured by the image (structure). Thus, in theory, you can take a whole bunch of pictures of say of a hillside or a glacier and create a digital surface model using SfM - intriguing!

Examples: Westoby et al. (2012)  or Ryan et al. (2014).
Nice review of SfM: Smith et al. (2016)

• Bundler (Univ. of Washington)
• VisualSFM (Univ. of Washington)
• Agisoft (commercial product)
• SFMToolkit

Fairbanks Fodar is a semi-commercial company out of Fairbanks (AK) claiming that they can deliver LiDAR quality data at 10 percent of the cost.