Research On Ship Abnormal Behavior Detection Technology Based On Multi-source Data Fusion

As the number of ships at sea continues to increase,it brings challenges to the monitoring and management of ships at sea.Effective detection and identification of abnormal ship behavior is of great significance for protecting marine safety and safeguarding marine rights and interests.AIS is the primary data source for target monitoring of ships at sea,but it is unable to monitor non-cooperative targets,while SAR can also detect non-cooperative targets.This thesis studies the detection technology of abnormal ship behavior by means of multi-source data fusion.The main research contents are as follows:1)Aiming at the problem of point trace association between SAR and AIS ship target,by extracting the geographical location information and geometric information of the ship target in the SAR image,the data association is carried out by means of multi-attribute combination.Considering that the nearest neighbor association algorithm based on location features is prone to mis-association in dense environments,a multi-attribute global optimal association algorithm is proposed,which take full advantage of the feature attributes of ship targets in SAR and AIS,and improves the accuracy of ship target point trace association.Through experiments and analysis,the algorithm has a higher average accuracy rate than the nearest neighbor association algorithm in dense environments.2)Aiming at the problem of abnormal ship behavior detection,the results obtained by associating SAR and AIS are used to detect abnormal ship behavior,and analyze several types of ship abnormal behavior discrimination methods.The ship abnormal behavior detection model based on AHP and fuzzy comprehensive evaluation can analyze and evaluate the abnormal ship behavior more comprehensively.The application of the algorithm model is analyzed through the case of illegal fishing behavior of ships.3)Design the ship abnormal behavior detection system,separate the computing platform from the specific business plug-ins in a plug-in way,and finally analyze and study the ship targets suspected of abnormal AIS equipment in the fusion results of SAR and AIS through the test results.