The Computer Vision (CV) workshop was held on the 24th June 2022 at the South African Marine Science Symposium held in Durban on the 20th to 24th June 2022. The workshop’s main goal was to connect marine scientists interested in this field with computer vision engineers and programmers. Furthermore, the workshop aimed to (1) showcase existing computer vision efforts in Southern African Marine Science, (2) create awareness of the latest developments in the application of these technologies worldwide and discuss their potential application here in South Africa.

 

The hybrid workshop was led by the Department of Forestry, Fisheries and the Environment (DFFE) – Drs Sven Kerwath (DFFE), Toufiek Samaai, (DFFE) Charlene da Silva (DFFE) and Andrea Angel (BirdLife).

 

The event was made possible with support from WILDTRUST, Rainforest Trust and Oceans 5. Dr Angus Patterson of the SAIAB African Coelacanth Ecosystem Programme (ACEP) has also secured seed-funding for two projects to be selected in the next few months.

WORKSHOP CONTENT

OUTCOMES

Presentations given during the CV workshop clearly indicated that there is a lot of scope to increase efficiency of most marine research programmes in South Africa using this technology. Given the scope of the work presented, it was clear that a lot of progress has been made in various fields but there is no cohesion within the Marine Science Community in terms of sharing these ideas, applications, and technology.  There is a lot of technical expertise in terms of engineering (building of cameras, BRUVS etc), and programming (AI and CV) in South Africa and with funding and direction we can build a platform to jointly improve on CV in the future. It was also clear that much of the engineering required to assist with projects can be sourced in South Africa, and information is available freely to learn how to use the techniques.

 

The group collectively decided that to move forward it will be helpful to create a formal working group the Computer Vision Marine Science South Africa (COVIMSA) that would facilitate the field moving forward from arranging meetings, organizing funding opportunities, hackathons and sharing latest useful applications for CV. In addition, the results of the workshop will be published in a peer-reviewed journal with all interested individuals included. Lastly the details of the seed-money to be provided by the African Coelacanth Ecosystem Programme (ACEP) will be released shortly through the SANCOR platform.

Computer Vision is a field of artificial intelligence (AI) that enables computers and systems to process and derive meaningful information from digital images, videos, and other visual inputs, thereby significantly decreasing time spent manually analysing digital input. The field of computer vision is concerned with automatic extraction, analysis and understanding of useful information form a single image or a sequence of images through development of a theoretical and algorithmic basis to achieve automatic visual understanding. Computer vision has the potential to significantly accelerate southern Africa’s marine environmental, biological and fisheries observation and monitoring and analysis capability. It can revolutionise many cost- or otherwise labour-intensive tasks in marine science, conservation, and fisheries applications.

 

Computer Vision has wide ranging applications in marine science and management of the marine space, for example:

  • Automatically classify and identify catch and bycatch species on fishing vessels during fishing, sorting or offloading.
  • Quantify marine pinnipeds and marine birds on breeding colonies via arial counts
  • Automatically classify and identify marine animals according to taxonomic features (e.g. sponge and sea-cucumber spicules, fish otoliths/scales, shark denticles).
  • Automatically classify and quantify habitats and or species on underwater or aerial footage
  • Automatically identify marine related events (boats, fishers, whale-blows, bird activity, algal blooms, sharks) via aerial footage and fixed-point and / or motion-sensing cameras along the shore and at harbours and slipways
  • Individual identification of marine organisms through pattern recognition