Project

Return to Project Catalogue

Automated spillway surveillance for risk assessment at FishPass

Selective passage systems allow passage of desirable fish while blocking movement of undesirable fish. Replacement of fixed, riverine barriers with effective intelligent migratory barriers will increase connectivity of waterways and aid species specific fisheries management and restoration efforts. The overall effectiveness of a selective passage systems will depend on both the selection mechanism and channel as well other components at the selection site designed to prevent passage (e.g., low-head weir). A hazard analysis for unintended escapement and video surveillance for unintended passage provide a more complete characterization of selective passage systems. To these ends, the launch angle and speed of steelhead (Oncorhynchus mykiss) present in the Laurentian Great Lakes were characterized through annotated images collected during spring 2022 and 2023 migratory runs. Analysis of 173 total leaps yielded a mean normalized launch speed of 8.6 body lengths per second and mean launch angles of 58.73 and 68.2 degrees in 2022 and 2023, respectively. To detect unintended passage over barriers, object detection methods based in machine learning were developed and evaluated. The developed models were trained to locate jumping fish and designed for incorporation with deployable network video recorders.

Status
Complete

GLFC ID
2022_EIC_793012

Research Program
FishPass

Research Theme
NA

Start Date
2022

End Date
2024

PI Name
Eickholt, Jesse

PI Email
eickh1jl@cmich.edu

PI Institution
Central Michigan University

Completion Report

Project Keywords



Project Datasets

Images of Leaping Steelhead for Launch Analysis

View Dataset

This dataset contains images related to 173 annotated steelhead leaps during migratory runs in the spring of 2022 and 2023. The images stem from a video camera positioned downstream and perpendicular to a weir on the Boardman (Ottaway) River, Traverse City, MI, USA. Through labeling of key markers contained in the images, the launch speed and launch angle of each jump is calculated.

Dataset Format
Compressed images (*.zip) and python notebook (*.i

Waterbody
Basin-wide

Variables
launch speed, launch angle, length

Dataset Keywords

Labeled Images of Emerged Salmonids in a Riverine Environment

View Dataset

The datasets contain over 2,300 annotated images of emerged salmonids collected in a natural riverine environment. The images stem from surveillance video collected during fall runs of steelhead (Oncorhynchus mykiss) on the Boardman (Ottaway) River in Traverse City, MI, USA. In addition to images of fully emerged salmonids, datasets are provided containing images of partially emerged salmonids, fully submerged fish, and other wildlife present in a riverine environment. The images are accompanied by annotations locating emerged fish.

Dataset Format
Images (*.jpg) and annotations (*.txt)

Waterbody
Basin-wide

Variables

Dataset Keywords



Other Project Products