Utilização de imagens spot 5 na detecção de embarcações marítimas brasileiras como suporte ao programa de rastreamento de embarcações por satélites

Abstract

Brazil, since 2006, has set up its “Programa de Rastreamento de Embarcações – PREPS” which is a system based on a Vessel Monitoring System - VLS to monitor its fishing vessels. The PREPS program relies on a ship-born component that provides the authorities with a continuous monitoring of vessels location and movements in real time. However, many ships are not equipped with this system, in fact only 10% of the active fishing vessels are equipped with it, the others 90% are not “visible” to the system. In this case remote sensing using Earth Observation satellites is recognized as an important tool to obtain information about these vessels and support the active system (VMS) with passive measurements for non- cooperating ships. Dra. Christine Corbane, 2008, has developed and tested within the framework of the LIMES project (Land and Sea Integrated Monitoring for European Security), an algorithm for automatic ship detection using SPOT-5 HRG data to complement existing fishery control measures, in particular the Vessel Monitoring System. The algorithm focused on feature-based analysis of satellite imagery. Genetic algorithms and Neural Networks were used to deal with the feature-borne information. This algorithm was designed to classify small targets such as shrimp boats and was successfully tested under French Guyana fishing conditions using panchromatic SPOT-5, 5-m resolution product. In this work we tested the “ShipDetection” in two regions of Brazil: the first one in North area quite similar to the conditions at French Guyana where the shrimp xiv fishery takes place and the second one at the South of Brazil taking into account the environmental and fishing context of its regions. The results showed that the algorithm works pretty well to detect small boats as the ones used in shrimp fishery at the North of Brazil and the longer ones used in the fisheries in the South of Brazil. Although some problems related to increased false alarm rates has to be solved particularly with images with a high percentage of cloud cover and sea cluttered background. Some more research is needed to turn the “ShipDetection” algorithm into an operational tool to monitor “Illegal and Unreported Fisheries” as detailed in this work.