Underwater Target Detection Software Demonstration on the RivGen Turbine

Publicly accessible License 

This repository contains data and processing scripts necessary to train the object detection models utilized in the underwater target detection software demonstration on the RivGen turbine project and to produce performance metrics (precision, recall, mAP50, mAP50-95).

- Contents -
Data consist of "images" and "labels". Each image has an associated label, both share the same time string in its file name (e.g., 2024_05_25_09_01_57.98.jpg and 2024_05_25_09_01_57.98.txt). Time strings have the format %yyyy_%mm_%dd_%HH_%MM_%SS.%3f. Images and labels were curated from 2021 and 2024 smolt outmigration periods at the project site in Igiugig, AK.

Images are monochrome 8-bit images of objects (smolt, debris, and other) passing through the field of view of the deployed cameras during various operational stages of the RivGen turbine.

Labels are text files indicating the class and bounding polygon of each object in an image. The provided labels use the "YOLO" label format.

- Requirements -
Python3.8+ is required to install and run the train and validation script.
The README.md provides instruction for installing the requirements from the requirements.py file.

- Instructions -
The "example_train.py" file ingests the provided data, trains a model, and produces model performance metrics at completion. NOTE: model performance metrics will vary from run to run as a consequence of the random selection of training and validation data.

Citation Formats

TY - DATA AB - This repository contains data and processing scripts necessary to train the object detection models utilized in the underwater target detection software demonstration on the RivGen turbine project and to produce performance metrics (precision, recall, mAP50, mAP50-95). - Contents - Data consist of "images" and "labels". Each image has an associated label, both share the same time string in its file name (e.g., 2024_05_25_09_01_57.98.jpg and 2024_05_25_09_01_57.98.txt). Time strings have the format %yyyy_%mm_%dd_%HH_%MM_%SS.%3f. Images and labels were curated from 2021 and 2024 smolt outmigration periods at the project site in Igiugig, AK. Images are monochrome 8-bit images of objects (smolt, debris, and other) passing through the field of view of the deployed cameras during various operational stages of the RivGen turbine. Labels are text files indicating the class and bounding polygon of each object in an image. The provided labels use the "YOLO" label format. - Requirements - Python3.8+ is required to install and run the train and validation script. The README.md provides instruction for installing the requirements from the requirements.py file. - Instructions - The "example_train.py" file ingests the provided data, trains a model, and produces model performance metrics at completion. NOTE: model performance metrics will vary from run to run as a consequence of the random selection of training and validation data. AU - Joslin, James A2 - Murphy, Paul A3 - Runyan, Alexa A4 - Scott, Mitchell DB - Marine and Hydrokinetic Data Repository DP - Open EI | National Renewable Energy Laboratory DO - 10.15473/2488381 KW - MHK KW - Marine KW - Hydrokinetic KW - ML KW - Machine Learning KW - Igiugig KW - Alaska KW - Detection KW - Cross-flow KW - Tracking KW - Smolt KW - Salmon KW - data KW - processing scripts KW - Python KW - perfomance metrics KW - TEAMER KW - underwater target detection KW - images LA - English DA - 2024/12/17 PY - 2024 PB - MarineSitu T1 - Underwater Target Detection Software Demonstration on the RivGen Turbine UR - https://doi.org/10.15473/2488381 ER -
Export Citation to RIS
Joslin, James, et al. Underwater Target Detection Software Demonstration on the RivGen Turbine. MarineSitu, 17 December, 2024, Marine and Hydrokinetic Data Repository. https://doi.org/10.15473/2488381.
Joslin, J., Murphy, P., Runyan, A., & Scott, M. (2024). Underwater Target Detection Software Demonstration on the RivGen Turbine. [Data set]. Marine and Hydrokinetic Data Repository. MarineSitu. https://doi.org/10.15473/2488381
Joslin, James, Paul Murphy, Alexa Runyan, and Mitchell Scott. Underwater Target Detection Software Demonstration on the RivGen Turbine. MarineSitu, December, 17, 2024. Distributed by Marine and Hydrokinetic Data Repository. https://doi.org/10.15473/2488381
@misc{MHKDR_Dataset_588, title = {Underwater Target Detection Software Demonstration on the RivGen Turbine}, author = {Joslin, James and Murphy, Paul and Runyan, Alexa and Scott, Mitchell}, abstractNote = {This repository contains data and processing scripts necessary to train the object detection models utilized in the underwater target detection software demonstration on the RivGen turbine project and to produce performance metrics (precision, recall, mAP50, mAP50-95).

- Contents -
Data consist of "images" and "labels". Each image has an associated label, both share the same time string in its file name (e.g., 2024_05_25_09_01_57.98.jpg and 2024_05_25_09_01_57.98.txt). Time strings have the format %yyyy_%mm_%dd_%HH_%MM_%SS.%3f. Images and labels were curated from 2021 and 2024 smolt outmigration periods at the project site in Igiugig, AK.

Images are monochrome 8-bit images of objects (smolt, debris, and other) passing through the field of view of the deployed cameras during various operational stages of the RivGen turbine.

Labels are text files indicating the class and bounding polygon of each object in an image. The provided labels use the "YOLO" label format.

- Requirements -
Python3.8+ is required to install and run the train and validation script.
The README.md provides instruction for installing the requirements from the requirements.py file.

- Instructions -
The "example_train.py" file ingests the provided data, trains a model, and produces model performance metrics at completion. NOTE: model performance metrics will vary from run to run as a consequence of the random selection of training and validation data.
}, url = {https://mhkdr.openei.org/submissions/588}, year = {2024}, howpublished = {Marine and Hydrokinetic Data Repository, MarineSitu, https://doi.org/10.15473/2488381}, note = {Accessed: 2025-04-24}, doi = {10.15473/2488381} }
https://dx.doi.org/10.15473/2488381

Details

Data from Dec 17, 2024

Last updated Jan 3, 2025

Submitted Dec 17, 2024

Organization

MarineSitu

Contact

James Josline

360.477.2901

Authors

James Joslin

MarineSitu

Paul Murphy

MarineSitu

Alexa Runyan

MarineSitu

Mitchell Scott

MarineSitu

DOE Project Details

Project Name Testing Expertise and Access for Marine Energy Research

Project Lead Lauren Ruedy

Project Number EE0008895

Share

Submission Downloads