Illegal Fishing Detection using Neural Network

Abstract

Illegal fishing has become a worldwide concern resulting in drastic ecological consequences due to activities like overfishing. It is statistically shown that about 11–20 million tonnes of fish have been caught illegally on an annual basis, which amounts to 14%–33% of the global annual fishing catch. The estimated illegal fishing catch is totaled to be around $23 Billion. The vessel's ability to dredge, deplete and damage has lowered the fish stock to 65.8% in 2017 from 90% in 1990 within the biologically sustainable levels. To serve the preservation of biodiversity, illegal fishing detection provides an inclusive analysis strategy on the available data from the automatic identification system (AIS), the relative position of a vessel could be identified and the radar detection aids the tracking of vessels. The data is gathered by satellites and terrestrial receivers which is analyzed by The Global Fishing Watch (GFW) organization. The model based on AIS data, speed of the vessel, and vessel type is used to predict the fishing status of a vessel. The model processes the data being fed and targets the vessel by behavior identification and the likelihood of illegal activity could be monitored.