2. Project Information:
The goal of this project is to develop a computer vision-based system to operationally estimate streamflow by leveraging USGS Hydrologic Imagery Visualization and Information System (HIVIS) (https://apps.usgs.gov/hivis/). The method consists of using the real time feed from a network of cameras deployed and maintained by USGS, perform an automated segmentation of the received frames, detect changes, and infer water level, water velocity,and discharge through a cross-correlation analysis of sequential images. The inferred quantitative information will enrich USGS observation record and support the Calibration/Verification of the NWM.
3. Project Description:
We are developing a software for this project to support data ingestion, and processing. To date, we have developed a computer-vision framework using python to classify and segment river ice from HIVIS images. This software report the percentage of ice coverage which will help the USGS to flag the ice-affected measurements.
4. Resource Requirements:
This software will be used by the USGS to report the percentage of ice coverage at two stations. However, this will be expanded to cover most of the HIVIS cameras. For now, we need an instance that has GPU capabilities. We are investigating the way we will communicate our results with the USGS. It will be either through a shared storage which USGS can access it 24/7, or by creating an API that they can use to get the required results. Also, we will pull data from another the HIVIS AWS storage for processing. In the development phase, all the results will be pushed to our on prem computers for validation and testing.
Options:
Cloud Provider: AWS
Required Services in the Cloud:
List of AWS Services
EC2
S3 (not now, but maybe)
EBS
EFS
RDS (not now, but maybe)
DynamoDB
Lambda
EKS (Kubernetes Cluster) (not now, but maybe)
5. Timeline:
This request is needed immediately and would extend until the project ends.If it continues to be funded into future years the resources needs will likely continue.
6. Security and Compliance Requirements:
None
7. Estimation:
We anticipate ~$500/mo. for EC2/EKS and ingress expenses described above. The cost may increase in the future based on the project expansion
8. Approval:
Once Form is submitted, we will email the CIROH management to get the approval.
Please use the existing CIROH's AWS Steven's subaccount for this and make sure to tag the resources with unique name so that we could monitor the cost. Thank you for your request. Let us know if anything.
1. Requester Information:
2. Project Information: The goal of this project is to develop a computer vision-based system to operationally estimate streamflow by leveraging USGS Hydrologic Imagery Visualization and Information System (HIVIS) (https://apps.usgs.gov/hivis/). The method consists of using the real time feed from a network of cameras deployed and maintained by USGS, perform an automated segmentation of the received frames, detect changes, and infer water level, water velocity,and discharge through a cross-correlation analysis of sequential images. The inferred quantitative information will enrich USGS observation record and support the Calibration/Verification of the NWM.
3. Project Description: We are developing a software for this project to support data ingestion, and processing. To date, we have developed a computer-vision framework using python to classify and segment river ice from HIVIS images. This software report the percentage of ice coverage which will help the USGS to flag the ice-affected measurements.
4. Resource Requirements: This software will be used by the USGS to report the percentage of ice coverage at two stations. However, this will be expanded to cover most of the HIVIS cameras. For now, we need an instance that has GPU capabilities. We are investigating the way we will communicate our results with the USGS. It will be either through a shared storage which USGS can access it 24/7, or by creating an API that they can use to get the required results. Also, we will pull data from another the HIVIS AWS storage for processing. In the development phase, all the results will be pushed to our on prem computers for validation and testing.
Options:
Cloud Provider: AWS
Required Services in the Cloud:
List of AWS Services
5. Timeline: This request is needed immediately and would extend until the project ends.If it continues to be funded into future years the resources needs will likely continue.
6. Security and Compliance Requirements: None
7. Estimation: We anticipate ~$500/mo. for EC2/EKS and ingress expenses described above. The cost may increase in the future based on the project expansion
8. Approval: Once Form is submitted, we will email the CIROH management to get the approval.