This session features presentations highlighting advancements in the collection, calibration, and processing of lidar and drone-acquired remote sensing datasets.
Methane Hotspot Mapping using Drones
Methane gas emissions is a hot topic that has important implications for many industries primarily due to increased scrutiny from local, federal and global government and environmental organizations. The current methods for detecting, quantifying and reporting emissions has limitations that can lead to frustration, conflict and a lack of confidence. In this presentation we will discuss the emerging technology of using UAVs or drones to detect and map methane emission hotspots occurring at large commercial and industrial sites. We will talk about the different types of sensors that are currently available, their strengths and weaknesses, and their overall applicability in the market. We’ll look at some of the lessons that we’ve learned along the way while developing a methane mapping product from scratch at a time when existing information and experience were non-existent in the marketplace. We will also look at where this technology fits into the emissions monitoring space, what value it can bring to industries and what it cannot do or is not a good fit for.
Caleb Cass, Firmatek
Using Principal Component Analysis to Improve Quality Reviews of Lidar Point Clouds
Lidar point clouds used to generate map products to ASPRS standards and for the Lidar Base Specification (LBS) for the 3D Elevation Program (3DEP) require rigorous quality reviews. Review procedures include automated checks and manual review (“eyes on”) in an interactive environment. Small site lidar surveys, such as drone lidar, require the same level of QA/QC review. Increasing the efficiency of the review process with confidence in the completeness of the review is an important topic for lidar data producers. We will present the results of our investigation into using Principal Component Analysis (PCA) to help streamline the QA/QC review, resulting in improved efficiency. PCA is used in computational geometry to characterize the spatial structure of surfaces. Dimensional characteristics of the local neighborhood, such as linearity, planarity, and sphericity can be derived. Used with additional metrics, such as the standard deviation along the surface normal (SDASN), these characteristics can then be used to automatically detect problem areas in a point cloud that require further investigation by a technician. In this presentation we will focus on the use of PCA to evaluate intraswath hard surface precision (smooth surface repeatability) of point clouds derived from airborne lidar as well as quantifying ground surface thickness for accuracy of classification reviews. Examples of the PCA approach as implemented in the LP360 software suite will be included.
Martin Flood, GeoCue Group
Towards the Automated/Semi-Automated Mapping of Geologic Faults via LiDAR Point-Cloud Data Processing
The Pajarito fault system (PFS), located on the Pajarito Plateau (PP) in north central New Mexico, is a seismically active region nestled within the Rio Grande rift. Understanding the number, extent, and location of the surficial expression of faults across the PFS has important implications for analysis of seismic hazards. However, ensuring completeness of lineament mapping across the PFS, as well as extension of mapping coverage to the North and South of the PP, is required to understand the PFS within the larger context of the Rio Grande rift system and its potential for future seismic activity. Given the large area of interest (~ 4mi x 7mi = 28sq. mi; 6.4km x 45km = 288sq. km), assistance with this fault mapping task via automated or semi-automated techniques as applied to lidar data is highly desirable. This presentation will illustrate our continuing efforts towards this goal. First, we describe processing of the lidar data to create polyline coverages of surficial fault candidates. Next, the use of these coverages along with the visual clues used by analysts to manually detect and trace lineaments, is explained, as well as the challenges we encountered. Finally, the calculation of accuracy assessment metrics, including modest but complimentary ground truthing efforts, are described.
Paul Pope, Los Alamos National Laboratory
Uncrewed LiDAR/Camera in-flight Calibration
This presentation is focused on LiDAR and camera in-flight calibration for uncrewed platforms. When using a LiDAR and a camera onboard a drone, boresight calibration becomes a necessity once the system is installed. LiDAR boresight calibration implies calculating the angular offset between the LiDAR frame of reference and the IMU frame of reference. Similarly, camera boresight is the orientation offset between the camera coordinate system and the IMU frame of reference. Camera interior orientation is correlated with camera boresight. Therefore, the UAV flight pattern as well as the data processing algorithm need to address that strong correlation. This presentation explains the best practice for LiDAR/Camera calibration which allows for obtaining accurate calibration parameters which guarantees producing accurate mapping products. Several data sets were acquired in the United States, Canada, and Germany to particularly assess the performance of the uncrewed LiDAR/Camera in-flight calibration mechanism discussed here. The results presented in this presentation prove that LiDAR and camera boresight calibration are necessary to be considered as part of the quality control strategy in an uncrewed LiDAR/Camera data processing workflow.
Dr. Mohamed Mostafa, Trimble Applanix
Implementing LiDAR in Aerial Surveying
This presentation explains in more detail UAS Aerial Data Acquisition and the best practices to acquire Survey Grade Accurate Aerial LiDAR Data. It describes a Survey Grade Aerial LiDAR Post-Processing Workflow, Residual Accuracy Assessments, etc. It begins on explaining the differences between Aerial LiDAR and Aerial Photogrammetry, and the Pros and Cons between the two. It then describes how we take that Aerial LiDAR Data from the field and process it correctly to obtain the levels of accuracy needed in the Surveying Industry. It dives into Aerial LiDAR Data Deliverables, Flight Parameters, etc. Overall, how to conduct and process and Aerial LiDAR Survey Grade Project from Field to Finish.
Austin Rains, Frontier Precision
Uncrewed Aircraft System (UAS) Data Calibration and Satellite Surface Reflectance Validation
To use remote sensing data acquired from UAS for studying land use change and management, the quality and provenance of the data must be well understood. This research presents the radiometric, geometric, and spatial processes developed at EROS that are necessary to validate the quality of UAS data. We present the results of investigation into radiometric targets (size and materials) that are recommended for generating consistent (with satellites such as Landsat and Sentinel) radiometric reflectance data from digital numbers recorded by the UAS sensors. We present the results from a frame-based sensor (Micasense ™) and pushbroom sensor (Headwall Nano ™). 26 targets at various reflectance levels were investigated and the measured reflectance were compared against the reflectance measured by the ASD Fieldspec™ spectroradiometer. The results are used to inform the “best” targets that could be deployed on field campaigns for converting raw digital numbers to reflectance using the two-point empirical line method (ELM) plot. We will also present implementation of the ELM method using Radcal One step software that has been developed at USGS EROS. For any organization placing high priority on the consistency and quality of data, the suggested validation steps important in ensuring that data are calibrated and their quality traceable via known processes and standards.
Mahesh Shrestha, KBR, Contractor to USGS
