(858) 717-0234 info@modusrobotics.com

LiDAR General Workflow

Collection Process
LiDAR Process
Analyze Process
LiDAR Collection Process
Flight Planning

Flight Planning

Here at Modus Robotics, we look at drone data collection as an industrial process. In remote sensing and data collection, you want a reliable and consistent process. As a pilot’s confidence and skill increase so do their awareness to use the best tools.

N

Check your airspace

Good operators make sure you obtain the right to fly.  This is getting clearance to fly near airports, notifying property owners, and getting municipal permits.

N

Determine Flight Profile for LiDAR and Mapping

Obtaining the best LiDAR Map is a function of speed, altitude, overlap, and precision flying.  Review our Point Cloud Planning Section.  You can do photogrammetry at the same time as LiDAR.  Depending on the camera you may have to fly lower and slower.

N

Load into Mission Planner

A mission planner allows you to observe the collecton flight path, determine how to manage changing terrain, and plan for obstacle avoidance

Fly

Here at Modus Robotics, we look at drone data collection as an industrial process. In remote sensing and data collection, you want a reliable and consistent process. As a pilot’s confidence and skill increase so do their awareness to use the best tools.

M

Don't Manually Fly

Beginning pilots tend to want to fly by hand. Flying manually:

  • Reduces battery life from unnecessary control inputs
  • Adds positional and heading errors that are hard to remove in post-processing
  • Increases collection time
  • Increases flight risks and human error
M

Don't use an app designed for photgrammetry

There are some really good autopilot applications for photogrammetry and we love them for photogrammetry. We have tried dozens of applications. However we find for LiDAR you need flight automation that allows you to change altitude, airspeed, and overlap independent of the camera.

U

What to look for in Autopilot System

Your Autopilot system should allow you to do the following:

  • Adjust flight over varying terrain
  • Load know obstructions
  • Load no-fly-zones
  • Load different elevation and terrain maps- especially if you are in the quarry industry.
  • Allow manual override in case of emergencies – “point and click operations”
N

Universal Ground Control Station (UgCS)

We recommend UgCS autopilot systems.  It is compatible with most mainstream flight controllers and has all these capabilities plus multi-drone simultaneous control and auto-continue, which allows you to plan a large area with multiple battery changes.   For many production drone pilots, the rates are defined by the kilometer or acre hence the more ground they cover the higher the profit.  UgCS is helping them become more profitable.

Monitor

N

Monitor Flight Operations

Once autopilot is engaged, the gear raises up, and the drone is moving to its first fly-to-point, it is time to monitor operations. If you are operating around sparsely populated people, on a job site, or near valuable infrastructure, anything can happen. This is why having reliable equipment and good planning is critical for good LiDAR collection.  A good autopilot will allow you to monitor flight and still maintain situational awareness of your surroundings.

N

Monitor LiDAR Collection

Professionals know even on good day conditions change. Monitoring LiDAR point cloud density ensures you get the data quality your analysts needs to build a decision-making product. Sometimes you may miss a checklist item, forget to turn on your LiDAR,  or conditions like varying vegetation requires you to change your flight profile.  You only know by monitoring.  For the professional, getting this information while in flight and making a change is critical.

N

Ground Control Station

There are many ways to control a drone, cell phone, tablet, laptop, or server station.  A good rule of thumb is the more flight and sensor tasks you are required to do, the more people or automation you will need.    If you choose automation, it is wise to have a screen to monitor that task.  As an example, the image is a mock-up of a three-screen server rack we have used on our mobile stations.  There we set up our flight autopilot and telemetry controls (left), first-person view or inspection camera (center), and our LiDAR Real-Time Viewer (right).

LiDAR Data Process
Augmented GPS

The process phase translates received sensor data from collected relative spatial information to usable coordinated reference information for further analysis.

The first part of this phase is geo-correction using various methods of global navigation satellite system (GNSS) Augmentation. Depending on the systems the process may start while you are in flight just after being received on board or after the information is uploaded to a computer or the cloud for processing.

The final step of this phase is to take the corrected georeference information and metatag all the data points and denoise the pointcloud so analysis can be performed.  For photogrammetry this is done directly to each image and when referencing the point cloud to Ground Control Points (GCP).  For LiDAR this is done directly to the relative point cloud.

Real-Time GNSS Augmentation Methods

GPS Correction Services

Satellite-Based Augmentation System (SBAS)

This is a class of augmentation system that supports wide-area augmentation through the use of additional satellite-broadcast messages. These systems have multiple ground stations, located at accurately-surveyed points. The ground stations take measurements of one or more of the GNSS satellites,   the satellite signals, or other environmental factors which may impact the signal received. Using these measurements, error corrective messages are created and broadcast via the satellite constellation.  Several countries host these systems.

Regional Systems

You should check if your vendor offers this standard or as an upgrade.  We offer SBA integration standard with all our products. 

RTK Method

Real Time Kinematic Method

Real-Time Kinematic (RTK) satellite navigation is a technique used to enhance the precision of position data derived from GNSS such as GPS, GLONASS. This method provides up to centimeter-level accuracy.

Components

  • Single base station receiver
  • Mobile units
  • Radio Modem

The base station re-broadcasts the phase of the carrier that it observes, and the mobile units compare their own phase measurements with the one received from the base station

This allows the units to calculate their relative position to within millimeters, although their absolute position is accurate only to the same accuracy as the computed position of the base station.

The typical nominal accuracy for these systems is 5 centimeter horizontally and 10 centimeters vertically.

Pros: This means a LiDAR operator would not have to geo-correct their data on the ground.  This is an advantage for firms that need to be corrected geospatial data really fast.

Cons: If you are flying longer distances, closer to the earth like most mapping drones do, then you are likely to lose the carrier signal.

Post-Processing GNSS Augmentation Methods

Post-Processing GNSS

Continuously Operating Reference Stations (CORS)

This is a distributed network of GNSS stations across the US.  The National Geodetic Survey (NGS) manages the CORS network. The service provides Global Navigation Satellite System (GNSS) data consisting of carrier phase and code range measurements in support of three-dimensional positioning and geophysical applications throughout the United States, territories, and a few foreign countries.  Surveyors, GIS users, engineers, etc use CORS to improve the precision of their positions. CORS enhanced post-processed coordinates approach a few centimeters relative to the National Spatial Reference System.

Pros: if you are within 20km of a site this is an affordable way to geo-correct your data without extra equipment.  it is also good to have as a backup for your GNSS system.

Con: These stations may not be close to remote working areas.

Click here for more CORS Info

Post Processing Kinematic

Post Processing Kinematic (PPK)

Like CORS, Post processed kinematic (PPK) is a GPS is a position location process whereby signals received from a mobile location receiving device stores position data that can be adjusted using corrections from a reference station after the data has collected.  The difference is this is equipment you set up near your collection site.

Components

  • Single base station receiver
  • Mobile units

Both mobile unit and base station receive position separately.  In post-processing, both data sets are compared and the common positional error is removed.

If  CORS station is close, it can be used instead of a base station.

The typical nominal accuracy for these systems is 2 centimeter horizontally and 4 centimeters vertically.

Pros: Less complicated set up.  Most reliable solution.

Cons: The data must first be downloaded and then the corrections process.

Compiling

Compiling LiDAR

Compiling LiDAR

When talking about data, compiling is the act of assembling several sources of information with the same references into one product.

What is being pulled together? 

  • LiDAR Table
  • Raw GPS and Inertial information
  • GNSS differential correction data

This is everything goes into LiDAR data table that can be displayed at a point cloud.

There are several factors that go into the best practices of compiling data.  I am writing this training right now and am leaving this in here a placeholder for a link to data segmentation processing and best practices training.

We will also have separate training specific to GEOMMS products in our membership site coming (Mar2018).  This will address using LiDAR tool; when to process with the first or last pulse; and importance of setting the correct geo-reference system.  

Clean Data

As with any process, errors, called noise, enter into the data table.  There are several methods we will cover in separate training on how to ensure you have the data without outlying information such as ‘blank data’, errors from geo-correction, and having too many points.

The End Product

When your data is complete you should have either a LAZ file or a LAS 1.2 to 1.4 file.  In some use cases, the file can be converted to an ASCI file.

To download the American Society for Photogrammetry & Remote Sensing (ASPRS) standard is for a LAS file, click here.  

Results

Results of Data Collection
LiDAR Analyze process

After the post-flight process, the acquired data determines the precise elevation and geospatial location of features on the earth’s surface. With advancements like initial auto-classification, multiple intensity returns, and increased pulse repetition rates, Drone Aerial Lidar Scans are an effective method for creating accurate HD 3D topographical maps and surveys of terrain elements and human structures.

While there are dozens of products that can come from LiDAR, there are essential products that make the foundation for everything else.  These baseline products are what we cover in this section, and we later blogs will address derivatives of these products and their use cases.

Point Cloud

DEM

DTM

DSM

Break lines

CHM

Contours

Point Cloud

Point Cloud

The most basic LiDAR-based output is a point cloud.  It is a relatively simple collection of georeferenced elevation points.  That is, at each point in the “cloud” of data, the horizontal (latitude, longitude) and vertical (elevation) positions are known based on the physical location of the LiDAR sensor and the return times of the laser signals retrieved by the sensor.

This is a data table of measured “points.”  Each point has a set of values: x,y,z,i, p, c, r,g,b. x,y,z is the coordinate of the point, i is the intensity of the return, p is which pulse return (1st , 2nd, etc), c is the classification, r is red, g is green, b is for blue.  R,G,B is always blank on a pure LiDAR collection.

Digital Elevation Model (DEM)

A digital elevation model (DEM) is comprised of elevation data points distributed along a regularly spaced horizontal grid, i.e., an elevation raster, analogous to a digital photograph, but with elevation data in the “pixels” instead of color data.  In a DEM, elevation data represent the height of the terrain without regard for vegetation or built-up structures, i.e., “bare earth”.  In practice, some computer interpolation is generally required to produce the regularity of the DEM raster grid which can introduce error.  Moreover, the spacing of the raster grid, if sufficiently large, can fail to capture relatively small linear features, like roads and streams.  Nevertheless, a DEM is an extremely efficient format for both storage and manipulation of elevation data.  Because of their similarity in structure to digital images, many of the same file formats used for digital images can be used with DEMs.

Digital Elevation Model

Digital Elevation Model (DEM)

A digital elevation model (DEM) is comprised of elevation data points distributed along a regularly spaced horizontal grid, i.e., an elevation raster, analogous to a digital photograph, but with elevation data in the “pixels” instead of color data.  In a DEM, elevation data represent the height of the terrain without regard for vegetation or built-up structures, i.e., “bare earth”.  In practice, some computer interpolation is generally required to produce the regularity of the DEM raster grid which can introduce error.  Moreover, the spacing of the raster grid, if sufficiently large, can fail to capture relatively small linear features, like roads and streams.  Nevertheless, a DEM is an extremely efficient format for both storage and manipulation of elevation data.  Because of their similarity in structure to digital images, many of the same file formats used for digital images can be used with DEMs.

DTM

DTM

A digital terrain model (DTM) is structurally similar to a DEM.  However, where elevation data are found along regularly spaced intervals in a DEM, in a DTM the elevation data can be, and frequently are, irregularly spaced.  Thus, elevations at particular locations can be included in the model without interpolation regardless of proximity to other points.  In practice, that means that the density of points can be adjusted to accommodate features in the terrain.  That is, where the terrain is complex, and/or contains relatively small linear features, more data are taken; where the terrain is simple, the collection of fewer points is acceptable.  Moreover, DTMs often include additional data, called break lines, that describe abrupt changes in terrain, e.g., cliff faces.  As a result of the additional complexity of DTMs, introduced by the irregular spacing of points and the break lines, they are more resource intensive.  However, they are technically superior to DEMs and are particularly useful for engineering applications.  The additional complexity, specifically that introduced by the break lines, requires the use of point and polyline capable file formats.

Digital Surface Model (DSM)

A digital surface model (DSM) is structurally similar, depending on the horizontal spacing of the data, to a DEM or a DTM.  However, while both DEMs and DTMs disregard above ground features, a DSM includes elevation data for both vegetation and built-up structures.  Thus, a DSM is a “non-bare earth model”.  As with DEMs, if a DSM is gridded some interpolation is generally required and small features may be lost.  Alternatively, a non-gridded DSM, analogous in structure to a DTM, can compensate for terrain complexity and small features, but would also be more resource intensive than a gridded DSM.

Digital Surface Model

Digital Surface Model (DSM)

A digital surface model (DSM) is structurally similar, depending on the horizontal spacing of the data, to a DEM or a DTM.  However, while both DEMs and DTMs disregard above ground features, a DSM includes elevation data for both vegetation and built-up structures.  Thus, a DSM is a “non-bare earth model”.  As with DEMs, if a DSM is gridded some interpolation is generally required and small features may be lost.  Alternatively, a non-gridded DSM, analogous in structure to a DTM, can compensate for terrain complexity and small features, but would also be more resource intensive than a gridded DSM.

Break lines

Break lines

Breaklines represent a distinct/abrupt change in the terrain.  They are comprised of a series of horizontal vertices with elevation data attached.  Breaklines in a DTM are visualized as edges in a TIN; they are stored as lines with 3-dimensional vertices in CAD or GIS formats.  They typically have the same vertical accuracy requirements defined for point clouds, and the linear features to be collected must be stipulated to ensure collection of appropriate data.

Canopy Height Map (CHM)

Forest or tree canopy density and height are used as variables in a number of environmental evaluations.  These can range from biomass estimation and vegetation coverage to determining biodiversity. Canopy density is the elevation ratio of vegetation to ground. Canopy height measures the elevation of the tree or vegetation tops to the ground beneath. DSM is a CHM is only the vegetation with the ground or DEM/DTM removed.  Lidar can be used to determine both of these variables.

Canopy Height Map

Canopy Height Map (CHM)

Forest or tree canopy density and height are used as variables in a number of environmental evaluations.  These can range from biomass estimation and vegetation coverage to determining biodiversity. Canopy density is the elevation ratio of vegetation to ground. Canopy height measures the elevation of the tree or vegetation tops to the ground beneath. DSM is a CHM is only the vegetation with the ground or DEM/DTM removed.  Lidar can be used to determine both of these variables.

Contours

Contours

Contour maps display isolines of elevation, i.e., lines of common elevation.  For example, a 50’ contour map would display contour lines representing elevation changes of 50’ from one line to the lines higher or lower on either side of it.  Contour maps are a traditional means by which three-dimensional surfaces have been displayed on two-dimensional maps, e.g., USGS topographic maps.  These maps can be produced from DEMs, DTMs, and TINs.  While contour maps are excellent for the visual interpretation of elevation changes from tangible maps, they are supremely inferior for both display and analysis in digital formats.