Skip to content

oceanhackweek/ohw24_proj_micronekton_img_pipeline_au

Repository files navigation

ohw24_proj_micronekton_img_pipeline_au

Project Name

Micronekton Imagery Pipeline + AI

Collaborators

Name Location Role
Ryan Downie Australia Project Leader
Bowen Zhang Australia Contributor
Reg Dowse Australia Contributor
Candice Untiedt Australia Contributor

Background

The PLAOS platform used by CSIRO collects vertical profile data through the water column including physical, acoustic and imagery data.

PLAOS platform

Goals

  1. Interface with the marimba platform by developing a PLAOS pipeline to manage and process data. Marimba is a Python framework designed for efficient processing of FAIR (Findable, Accessible, Interoperable, and Reusable) scientific marine image datasets.

  2. Apply an an existing AI-ML model FathomNet/MBARI-midwater-supercategory-detector to imagery from the oblique camera in order to detect midwater faunal classes and develop a mosaic of images for the detected classes.

Datasets

  1. PLAOS data: acoustic, imagery (vertical stereo cameras and oblique camera) and log data

  2. Oblique imagery from IN2020_V08

Results/Findings

  1. The FathomNet/MBARI-midwater-supercategory-detector was successfully run on oblique camera stills imagery from 15 stations (subset) to detect 16 faunal classes.

Outputs

a. Original image with bounding box predictions

Full size image with prediction

b. Cropped bounding box of predicted classes for each image

cropped bounding box

c. .txt file with all the bounding box information for each image

link to text file for this image

d. csv file with predictions and confidence levels for each image in the processed batch

link to csv file for processed images in this batch

e. mosaic of detected class images for each station, arranged by class and including a text overlay

Mosaic image for Station 07

About

No description, website, or topics provided.

Resources

Code of conduct

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Contributors 4

  •  
  •  
  •  
  •