Automated Monitoring

Building off our annotation tool and Squidle+, the project has also developed an automated monitoring tool. Using the the tagged images from the annotation tool, the automated monitoring tool is trained with AI to detect debris and species in ROV images. This helps to create faster and more accurate monitoring of underwater ecosystems. 

 

Automated Monitoring Tool Demonstration

Squidle+ is an online annotation platform for marine imagery that is also a centralized data management platform. It supports different annotation schemes tailored to the marine domain with the ability to create new ones. It is also possible to do the annotations collaboratively.

Background on the Use of Automated Monitoring in the Svalbard Demo Site

In addition to integrating the imagery from Adventfjorden with Squidle+, a demo notebook was created to illustrate an analysis for the georeferenced imagery dataset collected by the Blueye X3 in June 2023 from the Adventfjorden as part of the efforts to study the impact of untreated sewage in the fjord. It also has a demo of how AI models could be used to classify images with visual debris as well as estimating the area of debris/objects of interest in the images.

Image 4

Using Squidle+, Frequently Asked Questions

Accessing the Lab and BlueCloud

Instructional Video

Other practical considerations for using Squidle+

Images stored on Squidle+ must be georeferenced

Depth and altitude are recommended

Dataset owners are responsible for data storage and ensuring it meets requirements

Datasets are not hosted on Squidle+, but an external repository

Contact the Squidle+ team for assistance importing datasets

Squidle+ is domain specific and supports different label schemes, in which you can make your own

Serves as a centralised data management system where datasets can be citable and accessible in the future