Sea Cucumber Monitoring with AI and Machine Learning
Sea cucumber have been overfished by the modern seafood industry. This exploitation has sparked an interest in sea cucumber aquaculture, as well as the need for sea cucumber monitoring technology to detect and analyze their health and condition.
Marine Thinking’s artificial intelligence and machine learning have been trained to recognize and detect sea cucumbers on the ocean floor. The National Research Council of Canada (NRC) and the Atlantic Canada Opportunities Agency (ACOA) made this project possible with their support.
The team used remote operated vehicles (ROV) to navigate the ocean floor, collecting hours of optical footage across multiple trips.
Using portions of the captured footage, the team created several large data sets. These data sets manually identified and labelled sea cucumbers for the machine learning algorithm. The training process saw image augmentation techniques applied to optimize the learn quality. These techniques help us prevent data overfitting, increase prediction confidence, and improve the overall variability of the data models.
Once fully trained, the AI algorithm achieved successful sea cucumber monitoring during footage playback.
To visualize the data collected, the team produced maps for each fishing area. GPS coordinates and visualized location and density of the sea cucumbers living in each area populated each map.
If you have project that would benefit from our AI object recognition and tracking technology, we’d love to hear from you.
