Date post: | 20-May-2015 |
Category: |
Technology |
Upload: | wv-assocation-of-geospatial-professionals |
View: | 830 times |
Download: | 1 times |
Eastern West Virginia LiDAR Acquisition
Josh NovacProject Manager
Dewberry (Tampa, FL)[email protected]
Project Overview
Project Overview
• LiDAR Acquisition Scheduled for Winter/Spring 2012• LiDAR Acquisition is 100% Complete• Products are currently being delivered as they are completed• Collection designed to meet FEMA needs and USGS V13
Specifications for LiDAR
Project Overview- Specifications
• Nominal Pulse Spacing < 1 meter• Vertical Accuracy
– RMSEZ 12.5 cm– Fundamental Vertical Accuracy (FVA): 24.5 cm– Consolidated Vertical Accuracy (CVA): 36.3 cm– Supplemental Vertical Accuracy (SVA): 36.3 cm
• Relative Accuracy– Within an Individual Swath ≤ 7 cm– Between Swaths ≤ 10 cm
Project Overview- Specifications
• Spatial Reference System– Horizontal
• North American Datum of 1983• UTM Zone 18N• Meters
– Vertical• North American Vertical Datum of 1988• Geoid 2009• Meters
Project Overview – Specifications
• Breaklines– Inland Ponds and Lakes:
• 2 acres or greater• Flat and level (each vertex must have the same elevation)• Water surface must be at or just below adjacent ground
– Inland Streams and Rivers:• 100’ nominal width• Flat and level bank to bank• Should flow continuously downhill (monotonic)
Project Overview - Specifications
• LiDAR Classification– LAS format (v1.2) with ASPRS classification scheme
• Class 1 – Processed, Unclassified• Class 2 – Bare-Earth, Ground• Class 7 – Noise (High/Low Points)• Class 9 – Water (Classified Using Breaklines)• Class 10 – Ignored Ground (Breakline Proximity)
Project Overview – Deliverables
• Raw Point Cloud– LAS V1.2– Georeference Information in Header Files– GPS times recorded as Adjusted GPS Time– Intensity Values– Full Swaths– Size not to exceed 2GB per swath
Project Overview - Deliverables
• Classified Point Cloud– LAS V1.2– Meet V13 Specifications for Classification (The new V1 specs are now
out)– Tiled at 1500 m x 1500 m to U.S. National Grid
• Bare Earth Surface (Raster DEM)– Cell Size of 1 meter– ERDAS .IMG format (32-bit floating point)– Depressions/Sinks not filled (Hydro-flattened DEM not Hydro-enforced
DEM)
Project Overview - Deliverables
• Control– Supplemental Ground Control – Used to control the LiDAR collection
and processing– Ground Control Quality Checkpoints
• Minimum of 20 points across 5 land cover types– Bare Earth/ Open Terrain– Urban– Tall Weeds/Crops – Brush and Trees– Forested
• Must be on flat or uniformly sloping terrain
Project Overview - Deliverables
• Metadata– FGDC Compliant– Overview of processing steps and procedures
• Project Report– Detailed records of collection, production, and quality assurance
processes
Project Overview - Schedule
Deliverable Description Due Date Status
Mobilization 12/16/2012 Complete
LiDAR Acquisition 03/09/2012 Complete
Survey (QA/QC Points) 02/10/2012 Complete
LiDAR Calibration 05/11/2012 Complete
Pilot Deliverable 05/25/2012 Complete
Full Deliverable 11/15/2012 In Progress
Final Acceptance 12/15/2012
Project Overview - Contacts
• USGS State Liaison – Craig A. NeidigCharleston, WV
304-347-5130 [email protected]
• USGS Project Manager – Patrick EmmettRolla, MO
• Dewberry – Josh NovacTampa, FL
LiDAR Technology
What is LiDAR
• Light Detection and Ranging• Active Scanning System
– Uses its own energy source to produce pulses of laser (light) which are emitted, reflected and then received from surfaces
• Measures range distances– Based on time between emission, reflection and receive
time
• Direct terrain measurements, unlike photogrammetry which is inferred
• Day or night operation except when coupled with digital camera
• In addition to ranging, LiDAR systems can provide:– Additional information about the target (for classification)– Information about the transmission path (e.g. DIAL to
measure concentration of elements in the atmosphere)
What LiDAR is NOT
• The answer to all your elevation requirements• All-weather
– Target must be visible within the selected EM spectrum– No rain or fog– Must be below clouds
• Able to “penetrate vegetation”– LiDAR can penetrate openings in the vegetation cover but
cannot see through closed canopies
Airborne LiDAR System Components
LiDAR Transmitter, Scanner, and Receiver
Aircraft Positioning – Differential GPS (with post-processing)
Aircraft Attitude – Pitch, Roll, Yaw – Inertial Navigation System (GPS-Aided)
Data System
Operating Wavelengths
In theory, any light source can be used to create a LiDAR instrument Near-Infrared wavelength
Used by most airborne terrestrial LiDAR systems Easily absorbed at the water surface (unreliable water surface reflections). Wavelengths utilized: 1000 – 1500 nm
Blue-Green Wavelength Used by all airborne bathymetric and “topobathymetric” systems (532 nm) Can penetrate water, but signal strength attenuates exponentially through the
water column
GammaRays
VisibleUltraviolet Infrared Microwave TV/RadioX-Rays
100µm0.1cm
0.7Wavelength (not to scale)
0.30.2µm0.01µm 0.4 1.5 5.6µm 10cm1cm0.0001µm 1m100µm20µm
FilmElectro-optical Sensors
Thermal IR
Passive MicrowaveActive RADAR
Laser system characteristics
• Pulse width (or duration) is usually defined as the time during which the laser output pulse power remains continuously above half its maximum value (FWHM).
Pul
se w
idth
pulse width
“long” pulse
“short” pulse
time (ns)
inte
nsity
Multiple Scanning Patterns (two most common)
It is common to withhold the data for a few percent at the tips of the zig-zags where elevations are less accurate
Various LiDAR Formats
DigitizedBackscatterWaveform
Threshold
Discrete Return
Leading-Edge
PhotonCounting
Pulse-Width
Short DurationLaser Pulse
Image courtesy Dave Harding, NASA
Discrete return vs. waveform-resolving and the “dead zone” effect
Discrete-return LiDAR Waveform-resolving LiDAR most discrete-return systems require a minimum vertical object separation to
register consecutive returns from the pulse separately, thereby being blind to canopy material within this dead zone
Flight Planning Considerations
Maximum scan angle? Leaf-on or leaf-off?
Laser Penetration
Discrete Return LiDAR systems
All returns (16,664 pulses)
1st returns (11,460 pulses, 69%)
2nd returns (4,385 pulses, 26%)
3rd returns (736 pulses, 4%)
4th returns (83 pulses, <1%)
Image courtesy Hans-Erik Anderson
LiDAR Systems Manufacturers
• Leica Geosystems• Optech Inc.• Riegl
Enabling Technologies: Aircraft Position and Attitude Determination
Lidar System Components
• Lidar Transmitter, Scanner, and Receiver
• Aircraft Positioning – Differential GPS (with post-processing)
• Aircraft Attitude – Pitch, Roll, Yaw – Inertial Navigation System (GPS-Aided)
Differential GPS
• Positions the aircraft every one second
• Integrates with IMU – blends the GPS position fixes and the IMU orientation fixes for a complete record of the aircraft motion
• Largest source of positional error!
• Minimize baseline length for best accuracy
Inertial Measurement Unit - IMU
• Position and Orientation systems utilizes a combination of gyros and accelerometers
• Outputs position, roll, pitch and heading of airborne sensors real time and post processing
• Capable of 0.005˚pitch/roll, 0.008˚ heading accuracies (POS AV 510 post processed)
• 5-10 cm sensor positioning (post-processed)
• 200 Hz data rates
IMU - Orientation
Pitch Yaw Roll
LiDAR Data Processing
LiDAR Data Processing Workflow
DGPS Data
IMU Data
Lidar range
Scan Angles
Calibration and mounting
parameters
Point Cloud Data X, Y, Z data
Post-processed GPS trajectory and INS solutions
Data Processing Steps
• Initial processing done in field• Process GPS/IMU• Process calibration data• Process waveform data (if available)• Process full point cloud to calibration• Verify data (i.e. flight line comparison, coverage,
accuracy, etc.)• Post Processing – Classification; auto and manual
filtering
LiDAR: Raw Data Processing
• Data collected by flight• Monitored during collection
– Sensor operation– Flight line holidays– Data voids– Gross data errors
• Calibration flight at start and end of flight for adjustment of system and systematic drift
• GPS Data processing (kinematic post-processing aircraft GPS to reference station)
• Results in X Y Z, Scan Angle, Intensity, Return# ASCII or Binary files – Typically LAS
LiDAR post-processing creates a point cloud
LiDAR: Post Processing - Classification
• Separating ground from non-ground– Automated Processing– Manual Processing
Post Processing - Classification
• Automated scripts– Classifies approximately 80 – 85% and takes 20% of the time– Algorithm must be balanced to classify correctly - May cut into slopes too
much, or leave too much artifacts– Color coding orange = ground, green = other
Post Processing - Classification
• Manual Classification– Impossible to classify to the 100% level– Manual classification takes 80% of the post processing time (to get that last
20%)– Color coding orange = ground, green = other
ASPRS Standard LiDAR Point Classes
Classification Value (bits 0:4)
Meaning
0 Created, never classified
1 Unclassified
2 Ground
3 Low Vegetation
4 Medium Vegetation
5 High Vegetation
6 Building
7 Low Point (noise)
8 Model Key-point (mass point)
9 Water
10 Reserved for ASPRS Definition
11 Reserved for ASPRS Definition
12 Overlap Points
LAS Classified by Class
Elevation Data Challenges
• Large number of elevation records can require long processing times
• Exploitation of LiDAR has typically required specialized software such as• GeoCUE
• QT Modeler
• Terrascan/Terramodeler
• Many new LiDAR programs are being introduced which will allow more users access to the data
• ArcGIS – Version 10.1
• FugroViewer – Free
• LAS Reader for ArcGIS – Free
• PointView LE - Free
LiDAR Software Tools
• ArcGIS (10.1)• Geocue (Geocue)• LP 360 (GeoCue)• Quick Terrain Modeler (Applied Imagery)• Terrascan (Terrasolid)• LASTools• FugroViewer
Sample list – no endorsement is inferred or implied
Data Verification & Quality Control (QA/QC)
Data Verification & Quality
Three fundamental questions MUST BE ASKED
1. Did the LiDAR system work2. Are the data classified properly and free of
artifacts to support the intended product?3. Is the dataset complete?
Types of Analysis
• Quantitative Analysis– Utilize survey checkpoints to verify TIN accuracy– FEMA only “requires” quantitative analysis
• Qualitative Analysis– Subjective analysis to assess the quality which can include
cleanliness, usefulness for the intended product etc.• Completeness
– Are tiles complete with no voids, correct location, projection information, classified to the correct classes etc.
Dewberry’s Approach to QA/QC
Dewberry’s Approach to QA/QC
• Inventory (completeness)• Quantitative• Qualitative• Reporting
Quantitative Verification
• Ground truth surveys– Utilize GPS and conventional survey checkpoints (cp)– Place checkpoints in strategic locations based on flight line pattern– Verify data in varied land cover categories– Compare CP with interpolated TIN value
Qualitative Assessment - Techniques
• Utilize different software and tools • Use imagery• Create pseudo imagery• Combine images or techniques
• Bare-earth DTM• Full point cloud• Intensity Images• TIN’s• DEM’s• Hillshades• Contours• Slope Map• Speciality Maps• Survey Checkpoints/cross
sections
Derivative Products
Intensity Images
• Measures the amount of light returning to the sensor
• Useful for QA/QC & Research– Identify conditions at time of
collection
• Can be used for stereo-compilation to generate 3D breaklines (“LiDARgrammetry) or 2D features
Breaklines
• Linear features that control surface behavior• Can be 2D or 3D• Traditionally derived from stereo photogrammetry or from
surveys• Can use LiDAR and Intensity to create breaklines• 2D breaklines with assigned elevations for hydro-flattening are
typically used.
Terrain Dataset
A Terrain Datasets is a multi-resolution TIN-based surface build on-the-fly from feature classes stored in a feature dataset of a geodatabase.
Terrain Datasets are more effective for storing and visualizing large point data sets.
A Terrain Datasets resides in the same feature dataset where the feature classes (used to construct it) reside.
Terrain Datasets can be used to obtain TINs and grids.
Terrain Dataset
In a Terrain Dataset, feature classes include: Mass points (e.g., LiDAR); Breaklines (hard and soft); Clipping polygons (hard and soft); Erase polygons (hard and soft); Replace polygons (hard and soft).
A Terrain Dataset is composed of a series of TINs, each of which is used within a map-scale range. For each map-scale range, a level of detail (i.e., z resolution) and pyramid level are defined.
Different Treatments of LiDAR DTMs and DEMs
• Traditional Stereo DTM (Topographic Surface)
• Pure LiDAR (Topographic Surface)
• Hydro-Flattened (Topographic Surface)
• Full Breaklines (Topographic Surface)
• Hydro-Enforced (Hydrologic Surface)
• Hydro-Conditioned (Hydrologic Surface)
Traditional Stereo DTM (Topographic Surface)
• Reference image of the traditional stereo-compiled DTM
• Built from Masspoints and Breaklines
• Much coarser resolution than LiDAR
• Demonstrates the familiar and usually expected character of a topographic DEM
• Most notably, the “flat” water surfaces
Stream Waterbody
Pure LiDAR (Topographic Surface)
• DEM created only using bare-earth LiDAR points
• Surface contains extensive triangulation artifacts (“TINning”).
• Cause by the absence of:– LiDAR returns from water – Breakline constraints that
would define buildings, water, and other features (as in the Stereo DTM).
• Aesthetically and cartographically unacceptable to most users
TINning in Water Areas
Hydro-Flattened (Topographic Surface)
• The goal of the v13 Spec• Intent is to support the development of
a consistent, acceptable character within the NED
• Removes the most offensive pure LiDAR artifacts: those in the water.
– Constant elevation for waterbodies.– Wide streams and rivers are flattened
bank-to-bank and forced to flow downhill (monotonic).
• Carries ZERO implicit or explicit accuracy with regards to the represented water surface elevations – It is ONLY a cartographic/aesthetic enhancement.
• Building voids are not corrected due to high costs
• Most often achieved via the development and inclusion of hard breaklines.Stream Waterbody
Full Breaklines (Topographic Surface)
• A further possible refinement of the hydro- flattened surface
• Removes artifacts from building voids
• Refines the delineation of roads, single-line drainages, ridges, bridge crossings, etc.
• Requires the development of a large number of additional detailed breaklines
• A higher quality topographic surface, but significantly more expensive.
• Not cost effective for the NED.Buildings Roads
Hydro-Enforced (Hydrologic Surface)
• Surface used by engineers in Hydraulic and Hydrologic (H&H) modeling.
• Similar to Hydro-Flattened with the addition of Single Line Breaklines: Pipelines, Culverts, Underground Streams, etc…
• Terrain is then cut away at bridges and culverts to model drain connectivity
• Water Surface Elevations (WSEL) are often set to known values (surveyed or historical).
Culverts Cut Through Roads
Hydro-Conditioned (Hydrologic Surface)
• Another type of surface used by engineers for H&H modeling.
• Similar to the hydro- enforced surface, but with sinks filled
• Flow is continuous across the entire surface – no areas of unconnected internal drainage
• Often achieved via ArcHydro or ArcGIS Spatial Analyist
Common Data Upgrades to USGS V13 Spec.
1. Independent 3rd party QA/QC2. Higher Nominal Pulse Spacing (NPS)3. Increased Vertical Accuracy4. Full waveform or topo/bathy collection with red/green lasers5. Tide coordination, flood stage, plant growth cycle, shorelines6. Top-of-canopy (1st return) Digital Surface Model (DSM)7. More detailed LAS classification for vegetation, buildings8. Hydro enforced and/or hydro conditioned DEMs9. Single-line hydro feature breaklines; other breaklines10. Building footprints with elevations/heights11. Additional data products such as contours
Generating Contours from LiDAR
Contours are produced from LiDAR mass points and breaklines
Not aesthetically pleasing
ASPRS’ “DEM Users Manual”
1. Intro to DEMs, 3-D Surface Modeling, Tides
2. Vertical Datums3. Accuracy Standards4. National Elevation Dataset5. Photogrammetry6. IFSAR7. Topographic & Terrestrial Lidar8. Airborne Lidar Bathymetry9. Sonar10. Enabling Technologies11. DEM User Applications12. DEM Quality Assessment13. DEM User Requirements14. Lidar Processing & Software15. Sample Elevation Datasets
Final Report for NEEA Study available at www.dewberry.com
http://www.dewberry.com/Consultants/GeospatialMapping/FinalReport-NationalEnhancedElevationAssessment
THANK YOU
Josh NovacProject ManagerRemote Sensing Services LineDewberry (Tampa, FL)[email protected]: 813.421.8632