Modelling of Fish Swimming Patterns Using an Enhanced Object Tracking Algorithm

Modelling of fish swimming patterns has long been a challenge for the aquaculture community. In this proposed work, the modeling of fish swimming patterns and water condition such as pH, temperature and dissolved oxygen in fish tanks are analysed using network cameras and water sensors. An enhanced object tracking algorithm consisting of a combination of motion detection and condensation algorithm is applied to detect the change in speed and vector movement of the fishes in the viewing field. In addition, this research proposes an intelligent real-time monitoring tool to provide alert signals of water quality and healthy level of fishes.