Description
The dataset contains the data from the field deployment of the Unobtrusive Multi-static Serial LiDAR Imager (UMSLI) system at PNNL-Sequim test site in 2017. The dataset contains scanning of manmade models of baracuda and turtles as well as technical target placed in the field of view of 6 sets of LiDAR transmitter and receiver pairs. In addition, the system also operated to capture any marine life may come within the field of view. This included several encounters of Harbor Seal. In total, about 5TB of LiDAR data was recorded.
UMSLI is a LiDAR system uses lower power red (638nm, 180mW) laser that is invisible to the the marine animals. The red laser beams are scanned out of three bidirectional transmitters reflecting signals back to the receiver as it hits particles in the water, when a laser beam reflects as it hits marine life it is called a hard target. The main purpose of the system is to monitor MHK sites to detect and alert the site manager of any potential encounter between the energy generation equipment and protected marine life.
This submission includes a data lake resource for the raw data and experiment video recordings, MATLAB scripts for data parsing and analysis, and a folder of project publications. For more information on the data, see the "Data Description" resource below.
UMSLI is developed at Harbor Branch Oceanographic Institute at Florida Atlantic University under DOE funding (DE-EE0006787 and DE-EE0007828).
Resources
| Name | Format | Description | Link |
|---|---|---|---|
| 2. UMSLI Data Lake on AWS | HTML | Link for the UMSLI data lake. Includes videos and raw data broken out by month. For more information on the data see the "Data Description" resource. | Open resource |
| Marine Animal Classification Using UMSLI in HBOI Optical Test Facility.pdf | Journal article detailing the marine animal classification methods developed in the UMSLI project. | Open resource | |
| Marine Animal Classification With Correntropy Loss-Based Multiview Learning.pdf | Journal Article detailing the multiview learning classification method used in the UMSLI project and its importance. | Open resource | |
| MATLAB Scripts Updated.zip | ZIP | Code to parse and analyze data using National Instruments native library which significantly increases parsing speed. | Open resource |
| Publications.zip | ZIP | Folder containing conference papers and technical reports from the UMSLI project. | Open resource |
| Publications.zip | ZIP | Archive of publications generated during the projects. These included project progress reports, final reports, conference presentations, and journal articles. | Open resource |
| 1. Data Description.docx | Doc | Description of the data lake data and directory structure. | Open resource |
| MATLAB Scripts and Photos.zip | ZIP | Folder containing MATLAB scripts to parse and analyze data, as well as photos captured during experiment for analysis. | Open resource |
| Underwater LiDAR Image Enhancement Using a GAN Based Machine Learning Technique.pdf | Journal article detailing the GAN based machine learning technique used for LiDAR image enhancement in the UMSLI project. | Open resource | |
| Marine Energy Data Registry on AWS | HTML | AWS public dataset program registry page for data released under the Department of Energy's Water Power Technologies Office (DOE WPTO) Marine Energy Data Lake. The registry page contains information about dataset documentation, access, and contact, for each of the Marine Energy Data Lake datasets. | Open resource |