3D Mapping Limitations

by with 0 Comments

Mapping Failure

Failure to Resolve Shrubs and Trees

Three-Dimensional (3D) mapping with an aerial drone is exciting technology, but certain types of terrain will limit’s its usefulness. We’ll explain why some map jobs turn out good, others not so well.

Terrain Features that are Difficult to Resolve

In addition to our mapping service’s recommendation for photo set collection, we’ve found that terrain can be a significant limiting factor in the success of a mapping mission. Namely, shrubs and trees.

Many of our clients prefer terrain maps in the winter when the leaves are down so we can get ground-level elevation information. However, when the ground is covered with shrubs and trees the map doesn’t resolve very well. This is because the vegetation is too complex for the 3D processor. Remember, it’s trying to triangulate each pixel to assign its point in space. We’ve seen this even with overlaps as high as 90%/90%.

Notice the forested areas that don’t resolve well in this photo. You can see the trees have unfocused swirl artifacts. In more challenging areas, the map processor gives up and leaves the area blank. Although this picture came from our mapping service (which uses high-powered image processing), our panorama software PT Gui Pro also failed to resolve.

Other terrains such as developed properties have enough order to their colors, shades, and textures, that the map processor can triangulate the pixels. We’ve had excellent results mapping buildings, developed land, and roads.

What does our Mapping Service Recommend?

Our service includes these requirements for the photo set:

• Minimize areas with surface water

• Use an overlap of 75%/75% or better (we use 85 to 90%)

• Minimize windy conditions and long shadows

• Fly mapping missions at least 1.5 hours off solar noon

• The drone’s altitude should be at least 4 times the height of the tallest feature

However, our mapping service doesn’t have a specification for forested areas. We’ve talked about this, and they only go so far as to admit that these areas can be “challenging.”

Conclusion

We guarantee our map results and will not charge for results that our clients can’t use. However, mapping forested areas is risky for us so we may decline jobs like this. We’ll view your proposed map site with Google Maps and advise if it’s worthwhile to fly a mapping mission.

Altimeter Error’s Effect on 3D Mapping

by with 0 Comments

Aerial Drone Altimeter Error

Aerial Drone Altimeter Error

One of the specifications you won’t find for your aerial drone is its altitude accuracy. In fact, if you call the manufacturer’s support team, they probably won’t provide you with that spec. So, what do you tell a client when their mapping accuracy requirement is critical?

At FAD-Photo, we have calculated what you can expect in terms of altitude accuracy and can show you how this altitude error affects 3D mapping.

Your drone’s altimeter takes its readings from an internal barometer, does the conversions, and writes them to the photo tags, which are used by the mapping service. Mapping service providers will tell you that their numerical processing is accurate, but subject to the altitude information provided by the drone. But what if the drone’s altitude data is drifting over the duration of the mapping session?

The map processing service assumes the drone flies its mapping session in a perfectly flat plane. If the drone’s altitude readings shift during flight, then its plane shifts. This in turn shifts the ground plane causing terrain elevations to shift.

How does an Altimeter Work?

This blog is an update to our April 2020 blog Aerial Drone Altimeter Accuracy Specification. We analyzed nine months of data to refine our opinion on altimeter accuracy.

As one goes up in altitude, the air pressure goes down. The relationship between air pressure and altitude is quite predictable. So, using the formula below the altitude is easily converted from the drone’s internal air pressure sensor (a miniature barometer):

A = -(RT/gM)*ln(Po/P)

Where A is the drone’s altitude (height difference between the measurement altitude and the starting altitude), R is the gas constant, T is temperature of the air, g is the acceleration due to gravity, M is the molar mass of air, Po is the atmospheric pressure at the starting altitude, and P is the atmospheric pressure at the drone’s measurement altitude.

This formula is coded into the drone’s software so the barometer’s pressure reading can be converted to altitude. The constants R, g and M do not change, but the variables Po and P do change and are used in the calculation. The temperature T is assumed to be constant, but in reality, it’s affected by the temperature of the drone’s barometer and thus contributes to altitude error.

Over the course of a flight session, not only does the drone’s internal barometer sense pressure but it also senses the temperature effects of ambient air and heat dissipation of the internal components.

What Does the Data Tell us?

Our Flight Logs provided the record of end-of-flight altitudes for flights longer than 10 minutes. The resulting graph appears above.

The data indicate that altitude error is positive during the cooler months and negative during the warmer months. We know that average seasonal temperatures follow a sinusoidal curve, starting about 46 days after the autumn equinox (or 55 days before 1/1). Also, temperature variations are dependent on region, weather patterns, and time of day. In our case, this data was taken in central Virginia.

Developing a Trend Line

Using MS Excel’s Solver add-in, we derived the sine wave “trend line”, which follows this formula:

Altitude Error (trend line) = 2.44 + 4.74*sin((Date + 70.2)/365)

The resulting curve represents the altitude error in feet that may be expected throughout the year. The average offset is 2.44 ft plus a sinusoidal component with a peak value of 4.74 ft. The date offset of 70.2 days was expected to be closer to 55 days but is at least in the ballpark. As we collect more data, these numbers should become more accurate.

Statistical Data Calculated from the Trend Line

Other statistical data were calculated from the trend line minus the landing altitude data. We calculated a Standard Deviation of 4.60 ft and a Margin of Error of 1.14 ft (for a confidence interval of 95%). We then stacked these errors, so the expected altitude error at end of flight (EOF) is:

Altitude Error (EOF) = 2.44 +/-4.74 +/-4.60 +/-1.14 ft

Which ranges from 12.93 to -8.04 ft throughout the year

Since the drone altitude at takeoff is zero, we know this error has to build up over the course of a flight. Approximating this buildup as linear, the midpoint error is about half of the EOF altitude error:

Altitude Error (midpoint): ranges from 6.47 to -4.02 ft throughout the year

This means that for a 20-minute mapping session, the altitude error specification is around +/-6.5 ft. That is, the error can start off at zero and end at 13 ft, about a mean of 6.5 ft. The graph shows that the worst-case error can be slightly higher, but in most cases is somewhat less.

Conclusions

We take our maximum calculated error of +/-6.5 ft and round up to +/-7 ft, which we use as our altitude specification. Altitude inaccuracies affect topography mapping because our mapping service takes altimeter readings from the photo tags and cannot compensate for these errors.

For topography maps, contour lines at 10-ft intervals approach the useful limits of this technology. We can generate 5-ft contours (or less), but the overall mapping information becomes less and less useful. However, contour lines on this order may still be useful for assessing localized areas, such as slopes and structures.

Altitude error will be reduced for shorter duration mapping sessions and (as seen on the graph) near the sine wave crossings in late May and late September.

So, why won’t your drone manufacturer tell you its altimeter specification? Well, it’s complicated . . .

Seasonal Variations in Aerial Mapping

by with 0 Comments

Summer Map, 10-acre site

Summer Contour Map, 10-acre site

Seasonal variations are an important consideration for mapping your property or structure. For summer mapping, that is the late spring to early fall, the landscape is alive with all the vibrant colors that make for great mapping photography. Although the winter months are less colorful, there are significant advantages to these maps as well.

Summer Mapping

Summer mapping is ideal for showcasing properties and structures, especially for real estate sales. Overhead maps capture the properties with beautiful colors, not to mention stunning detail. (Think of Google Maps, but with super high resolution sufficient to see small objects, such as people and animals.)

All of our map products include geoposition and altitude information, so features such as a structure’s location and height can be measured. Our map products are referenced to sea level, in units of either feet or meters. Position and altitude information are available by just clicking the desired point.

The drawback to summer mapping is that vegetation and leaves hide the landscape that lies below.

Winter Mapping

Winter mapping is ideal for topographical charting of land features otherwise masked by vegetation and leaves. Our mapping software can “see” through naked trees and capture much more of the land features otherwise obscured in summer.

Developers of properties use our topographical products to design projects and estimate their costs. The three most common map products that aid in their decision making include:

  1. Contour Maps. We develop contour lines at the interval specified by our client. They can be at any interval, such as 50 feet, 25 feet, 10 feet, etc., and any unit, such as feet or meters. We have advanced post-processing techniques that we use to overlay contour maps onto our color maps. Examples of our composite maps are shown in these summer and winter pictures.
  2. 3D Object Maps. When opened in an object viewer, these maps provide the client with a look at the property from any angle (both above and below) the image. These full-color images provide height and perspective information of landscape and structures.
  3. Point Cloud Maps. These maps provide 3D views of the map image. They appear as a cloud of points, but each point has position, altitude and color information. The real power of these maps is their ability to see landscape underneath the trees and give the project engineer detailed information on features such as mounds, river banks, small structures, etc. Any particular map section can be selected and viewed. The selection can be rotated and zoomed to view the landscape features better than an in-person survey.

For more information on precision 3D mapping, please read our June 12, 2020 blog.

Mapping Challenges

Winter Map, 10-acre site

Winter Contour Map, same 10-acre site

Our aerial drones take overhead photos shooting straight down and in rectangular patterns. At a flight altitude of 400 feet, the ground resolution is typically 1.25 inches per pixel and the resulting map size is approximately 4 megapixels per acre. OK, this is some serious resolution!

However, there are certain areas that don’t resolve well in aerial maps. Water features and non-distinct land features may be difficult to resolve because discernable points cannot be identified or they’re in motion. These challenges are minimized with high overlap photography. That is, overlapping the photos at 90%. (This means taking 18 photos per acre.) Even at high overlap settings, there still may be features that don’t resolve well, such as bodies of water.

Why? Map making software identifies overlapping pixels to determine their exact position in space. At a 90% overlap setting, a single pixel may have as many as 100 look angles, where each angle helps to establish that pixel’s exact position. Errors in calculating these angles lead to errors in the map’s presentation.

Map processing generally goes well at 90% overlap, but can degrade at lower overlap settings, wind conditions, water features, and non-distinct land features. Winter mapping is usually more challenging because land features can be drab and non-distinct.

Which Season is Right for You?

We wrote this blog to take out some of the mystery of good map making techniques. At FAD-Photo, we have developed many photo maps and know how to set up your 3D map products regardless of season.

Aerial Drone Preflight Checks

by with 0 Comments

Checklist

Checklist

Preflight checks of your aerial drone are always a good idea, especially when you’re flying for a client. You may have seen a requirement from your insurance company to follow a written Standard Operating Procedure. That entails a checklist, which helps you to ensure your drone is ready for flight and to carry out its intended mission.

What does the FAA have to say?

The FAA’s position comes from the standpoint of safety of operation. Here’s the relevant sections:

14 CFR Section 107.15  Condition for safe operation:

(a) No person may operate a civil small unmanned aircraft system unless it is in a condition for safe operation. Prior to each flight, the remote pilot in command must check the small unmanned aircraft system to determine whether it is in a condition for safe operation.

(b) No person may continue flight of the small unmanned aircraft when he or she knows or has reason to know that the small unmanned aircraft system is no longer in a condition for safe operation.

FAA Advisory Circular 107-2 (June 2016, active to this date)

Para 5.9. Preflight Familiarization, Inspection, and Actions for Aircraft Operation. At 296 words, I’d rather provide you with a link than repeat it in this blog.

Para 7.3. Preflight Inspection. Ditto, 337 words, so please refer to the link.

Both paragraphs are well worth the five minutes to read.

What does your Insurance Company have to say?

Several insurance companies require aerial drone pilots to follow a Standard Operating Procedure (SOP). Although your drone manufacturer may not provide one, they’re fairly easy to develop, especially using the information in the FAA Advisory Circular.

Tips for Preflight Checks

Develop your own checklist.  We suggest including the following:

Software and firmware are up to date

SD card is installed

Camera lens is clean

Propellers are in good condition

Fresh batteries in your drone, controller, and cell phone/tablet

Develop a mission profile for your client and review prior to flight

Check the weather forecast, note the conditions prior to flight

Take off and hover at 5 feet; check propellers and flight controls

Add to this list as you see fit and go through it every time you fly. Pretty soon, your preflight checks will become second nature.

Precision 3-Dimensional Mapping

by with 3 Comments

Topography Map, May 2020

Topography Map Draped Over the 2-D Map

Precision 3-Dimensional Mapping

Aerial drones are the ideal method for collecting precision aerial mapping information for your land development projects. This is exciting technology and the map products that we deliver are truly breathtaking.

Each project begins with a client-provided map that outlines the site that needs to be surveyed. We enter this information into our drone’s autopilot (a mapping application), which flies the drone and collects the photos. Our typical settings are 90% overlap and 3 cm/pixel, which are further explained in our Orthomosaic Mapping and Photomapping blogs (parts 1 and 2).

We’re very good at photographing and delivering precision map products. As described below, several of these deliverables require specialized software to take full advantage of 3-D mapping. We do not offer professional cartography services, but instead provide these files to professionals who have the specialized software for these types of projects. The free software applications described below are suggested for viewing our products, but are not endorsed by FAD-Photo as suitable for professional-level mapping. We do believe, however, that many users will find them quite useful.

3-D Map Processing

Using our typical settings, the drone takes 18 photos per acre of land. For large sites, where we collect hundreds of photos, each pixel of the surface is examined at 13 or more different angles. Map processing aligns the pixels and assembles them into a 3-D composite model that includes latitude, longitude, and elevation.

Accuracy? Each photo is tagged with its position and altitude, so the composite model’s position is as accurate as the Global Positioning System. Typically, 3-4 meters.

Altitude information is based on the drone’s barometer, which has an accuracy of 3-4 meters. (We covered this specification in our April 23, 2020 blog.) Map deliverables are normalized to sea level.

Image processing is highly complex, so we use a professional mapping service provider. These are the deliverable products you will receive:

Full Color 2-Dimensional Map

This JPG file is a composite map of the photos, which are combined into a single panoramic map. Instead of a traditional scale, such as 10 meters per centimeter (or 100 feet per inch), the map service provides scale in terms of centimeters per pixel (or inches per pixel).

The JPG map doesn’t include position information, but its TIF counterpart (also a deliverable) has position information for each pixel. Use an application, such as the free QGIS software to view.

3-Dimensional Maps

DEM – Although monochrome, the Digital Elevation Model map (a TIF file) includes position and elevation information for each pixel. Special software, such as QGIS, must be used to view. The mapping service also provides a JPG of the DEM map, but this product doesn’t include position information.

Point Cloud – This is a LAS file, developed for LIDAR applications. At first glance, this full-color type of 3-D map appears fuzzy and not very useful. However, with a good viewer, such as the free Fugro Viewer, you can zoom in on the left panel image and view its corresponding 3-D model on the right panel. This is useful for looking at pixels under trees which would otherwise be masked. Of note, the 3-D model can be rotated in any direction with the mouse.

3-D Object Map – This is also a full-color map that can be rotated in any direction with the mouse. It offers a much sharper appearance than the point cloud, but it doesn’t get under the trees. Three files are required: the main 3D.OBJ file, a 3D.JPG file, and a 3D.MTL file. (You can rename the OBJ file, but don’t rename the other two.) You can open this type of map with the Windows 10 Object Viewer, but the free MeshLab viewer allows full 3-D rotation and zoom with the mouse.

Other Deliverable Map Products from FAD-Photo

The map processing report provides details on your map products, including map location, output size in pixels, scale in inches per pixel, overlap report, etc.

Topographical map (traditional contour map), where the user can specify the color scheme and contour intervals. (A postprocessing fee applies.)

Topographical map draped over the panoramic map. Here, the contour intervals are overlaid onto the full-color 2D map. An example is provided above and a larger example appears on our portfolio page. (A postprocessing fee applies.)

Do you have a special application?

Contact us for the solution. We’re experts in drone photography, mapping, and postprocessing services.

Orthomosaic Mapping and Photomapping, Part 2

by with 1 Comment

This 40-acre cornfield is a composite of 326 images

This 40-acre cornfield is a composite of 326 photos

Part 1 focused on orthomosaic photogrammetry maps – our high-tech mapping service that delivers image files with embedded position and altitude information for each pixel. Another service that we provide – photomapping – uses a more familiar image processing technique known as photo stitched panorama.

Although the photo collecting technique is similar to orthomosaic mapping, position and altitude information are tossed in favor of the less complicated panorama image processing. We use our licensed professional software to align features in the overlap areas, stitch the images together, and shade the transition zones.

Panorama Software is Common, but How Does it Work?

Panorama software works best when the camera is fixed and all images are taken from the same point in space. However, when using a moving camera the altitude must be high and a large overlap used. Therefore, we use the same flight control software and settings for photomapping that we use for orthomosaic mapping.

Our photomaps have the same high resolution as orthomosaic maps, but avoid the expense of the mapping service for generating 3D map sets. If position and altitude information aren’t required, then this is a less expensive way to get ultra-high-resolution maps of large properties.

Map Image Resolution (Geek Alert!)

Our Phantom 4 Professional V2 drone takes images that are 5,742 pixels across and 3,648 pixels high, which yield an image size of 19.98 megapixels (rounded to 20 MP). At an altitude of 400 feet, that image represents a ground view that’s 600 by 400 feet. Dividing pixels by distance yields the spatial resolution, which in this case is 9.6 pixels per foot. (In metric units, that’s 3.2 cm/pixel.)

For example, take an acre of land, which is 43,560 square feet. When photographed at an altitude of 400 feet, one acre takes up 18% of the camera’s image. The area-to-image ratio is scalable, so an orthomosaic map or photomap of 40 acres can be covered with 8 images. The composite image size is then 145 MP, with a resolution of 3.2 cm/pixel. The key word here is “composite,” since a large number of photos (e.g. 326 photos at 20 MP each) contribute to this composite image.

Since a map’s image size is proportional to the number of acres, we can estimate your finished map’s image file at 3.6 MP per acre. This number can grow 20% or more because we’ll always be photographing a larger tract than required. In terms of file size, the finished JPG image file will end up at about 1.5 MB per acre.

Want to Proceed? Here’s What We Need to Know:

When specifying a mapping job, clients just need to provide their tract boundaries and we’ll take care of the rest. We’ll determine the appropriate parameters, such as 85% overlap frame to frame, 85% overlap track to track, and flying altitude. Overlap is partly determined by the height of objects on the ground (such as trees) and seasonal variations (such as leaves).

Our flight control software will use this information to generate the photo-taking commands to be used by the drone’s autopilot, which ensures the photo-taking process is accurate and repeatable. This is especially useful if the map needs to be updated for project progress or for seasonal variations.

Viewing Large Image Files on Your Computer

For our example above, a 145 MP image is too large to display on common photo viewers such as the MS Office Picture manager. However, it can be viewed with more advanced software such as Adobe’s Photoshop. Of note, we can resize large images so they’ll display on your photo viewer, but the resolution will have to be decreased.

How Much Will Map Services Cost?

Every job has its own unique requirements, so we don’t publish our prices. However, our prices are very competitive and we deliver an excellent value. We’re happy to take your map requirements and give you an estimate within 24 hours. We guarantee our work will meet or exceed your requirements.

Call or e-mail. We’d love to hear from you!

Orthomosaic Mapping and Photomapping, Part 1

by with 2 Comments

Aerial Drone Photography is the Ideal Technology for Orthomosaic Mapping and Photomapping!

Digital Elevation Model

Digital Elevation Model Topographical Map

Realm of Possibilities

In this article, the term orthomosaic mapping is used to describe the orthomosaic photogrammetry mapping technique, which is a computationally-intense method that yields position and altitude information for each pixel in the map. The term photomapping is used to describe the use of photostitching software to generate large panoramic maps.

Our orthomosaic map deliverables include: ultra-high resolution 2D maps, 3D Digital Elevation Model (DEM) topographical maps, 3D models, and 3D point clouds.

Our Photomap deliverable is a 2D map image similar to Earth-type maps but with ultra-high resolution.

As implied, “high-resolution” means these map image files can be very large – on the order of 3.6 megapixels per acre. For example, an 80-acre map will have around 300 MP and a JPG image file of around 120 MB. Large files for sure, but the terrain detail is amazing!

Who Can Use Orthomosaic Maps?

Surveyors, Architects, and Civil Engineers are several of the many professions that use orthomosaic maps for their land development and construction projects. They have the budget for high-end software, which can further process our orthomosaic map products. Tract size can range from less than one acre to thousands of acres.

Real Estate marketers and landowners may also need high resolution maps but don’t want to make the significant investment in photogrammetry software. For these users, we offer photomaps, which provide beautiful full-color map images. More on photomapping in Part 2.

Either way, these maps have high enough resolution to detect very small features. For example, a 1 foot by 1 foot object is represented with around 100 pixels.

What Makes Aerial Drone Orthomosaic Maps So Special?

Three dimensional computing is at the heart of calculating position and altitude information for each pixel. The numerically-intense software that does this, to my knowledge, hasn’t been made available for personal computers, so most users have to use online mapping services. For more detailed information on how this specialized software works, please refer to this article at ScienceDirect.com.

How Is It Done?

Orthomosaic mapping software requires many photographs of the landscape so each pixel gets multiple look angles. The software then assigns position and altitude information to each pixel. As you might reason, this is a very complicated process. To get good results, very high overlaps of the subject area are required – on the order of 75 to 90%. This requires that each ground point is photographed 16 to 100 times.

Aerial drones with precise GPS-based navigation are ideal for photographing landscape with this kind of precision. Hundreds or even thousands of high-resolution photographs are taken, typically looking straight down, and stored on the drone’s internal memory card. Back at the office, these photos are then uploaded to the mapping service and reconstructed by their orthomosaic software to create stunning, full-color maps, DEM maps, 3D models, and point clouds.

To achieve good results, the drone’s altitude must be around 4-5 times the altitude of the highest object, such as trees. The algorithms work best when there’s low wind and lots of leaves on the trees. Water can be a challenge due to its reflectivity. With less-than-ideal photography, the algorithms have difficulty assigning position and altitude information to the pixels. If done incorrectly, the resulting map either has strangely-shaped areas or areas that are blanked out.

Orthomosaic Map Deliverable Products:

  • 2D Map – full-color image of the landscape, including position and altitude information.
  • Digital Elevation Model – color-coded for elevation, any color scheme is possible as well as contour lines.
  • 3D model – full color three dimensional image. The model can be viewed on-screen from any perspective.
  • Point Cloud – typically viewed with high-end software. It’s a full color 3D model using points.

In part 2, we’ll examine photomapping, which will be of particular interest to real estate marketers and landowners.