Mar 25 2020
9:00 am - 10:00 am
Academic Research Programs
Track Names: ACADEMIC RESEARCH PROGRAMS, Wednesday 9:00 - 10:00
Session Date: Mar 25 2020 9:00 am - 10:00 am
NASA’s Earth Observing Laser Altimeter ICESat-2: Mission Status Update and Data Product Performance Evaluation
9:00 AM - 9:15 AM
The Ice, Cloud and Land Elevation Satellite-2 (ICESat-2) launched in September of 2018. The satellite carries a unique laser altimeter that provides global elevation measurements. Although the primary scientific goal of the mission focuses on the cryosphere the measurements are also actively supporting many other scientific disciplines that include terrestrial ecology, oceanography, hydrology/bathymetry and atmospheric applications. Since on-orbit the preliminary data has allowed for an understanding of the scope of contributions this mission will make to science. Validation efforts for the data have revealed a performance that meets or exceeds the mission requirements for accuracy and precision, an exciting result for a global observatory. This discussion will address the operational aspects, technology details and data products of the mission with an update on the current mission status and performance levels. The presentation will also take an in-depth look at validation of the land and vegetation product, bathymetry studies, and terrestrial ecology applications associated with using the ICESat-2 data independently and comprehensively using other remote sensing products in unique combinations.
Horizontal Calibration of Vessel Lever Arms Using Unmanned Aircraft Systems (UASs)
Knowledge of lever arm distances from sonars, mounted on vessels, to systems such as Inertial Measurement Units (IMUs) and Global Navigation Satellite Systems (GNSS) is crucial for accurate ocean mapping applications. Traditional methods, such as laser scanners or total stations, are used to determine professional survey vessel lever arm distances reliably. However, for vessels of opportunity that are collecting volunteer bathymetric data, it is beneficial to consider survey methods that are less time consuming, less expensive, and which do not involve bringing the vessel into a dry dock. With the development of Unmanned Aircraft Systems (UASs) in the field of mapping, more cost-effective and quicker surveys can be conducted. To investigate the feasibility of conducting accurate horizontal lever arm surveys of vessels, while maximizing time efficiency in data collection, UAS surveys of a vessel with calibrated lever arm distances were conducted using both Structure for Motion (SfM) photogrammetry and aerial LiDAR while the vessel was docked at the pier. Estimates of the horizontal uncertainties, for both methods, were obtained by comparing the horizontal distances between targets acquired by the UAS methods to ground-truth measurements of lever distances from survey-grade laser scanning of the vessel. With the use of Ground Control Points (GCPs), horizontal uncertainties of both the photogrammetry and LiDAR models are on the order of centimeters, with the LiDAR model being slightly higher in horizontal uncertainty than most of the photogrammetry models.
Center For Coastal And Ocean Mapping (CCOM)
Deep Learning-based Classification of Large-Scale 3D Point Clouds
LiDAR and image-based remote sensing technologies are widely used to capture entire landscapes, cities and infrastructure networks. Classification, interpretation and information extraction are essential processing steps for preparing the captured data for a variety of geo-spatial applications in areas such as urban planning, environmental monitoring, and disaster management. In this talk, we discuss the potential of artificial intelligence for geospatial applications and demonstrate how machine learning concepts and GPU architectures can be combined to efficiently and reliably classify large-scale 3D point clouds. We present deep learning techniques that can be applied to highly detailed 3D point clouds, allowing to detect arbitrary objects and structures within. We demonstrate the practicability of the proposed techniques based on several case studies using mobile mapping data from rail as well as road networks. Using a modular, configurable processing chain and a small training data set, the captured raw data was labeled as, e.g., markings, poles, signals, buildings, vegetation and ground. During the talk, we will give insights into the technologies used along this processing chain. The results show that a deep learning-based classification opens up new ways to automatically process and analyze large-scale 3D point clouds as required by a growing number of applications and systems. The modular processing approach allows to scale across different hardware setups.
Hasso Plattner Institute
Q&A and panel discussions with session presenters
There will be a 15 minute Q&A /panel discussion with the presenters of the Academic Research Programs session.