Identification algorithm of fishing vessel operation type based on Feature Fusion
| Autor: | Dan Liu, Zhengang Zhai, Bingtao Gao, Lei Wang, Yunya Zhu, Li Zhang, Yinjie Wang, Tengjun Yao |
|---|---|
| Rok vydání: | 2020 |
| Předmět: |
0209 industrial biotechnology
Feature fusion Computer science Fishing 02 engineering and technology Sensor fusion GeneralLiterature_MISCELLANEOUS Task (computing) Identification (information) 020901 industrial engineering & automation Feature (computer vision) 0202 electrical engineering electronic engineering information engineering 020201 artificial intelligence & image processing Algorithm ComputingMethodologies_COMPUTERGRAPHICS |
| Zdroj: | 2020 IEEE International Conference on Artificial Intelligence and Information Systems (ICAIIS). |
| Popis: | In order to make full use of the information contained in the data, improve the accuracy of identification and classification of fishing vessel operation type, this paper transformed the problem of the identification for fishing vessel operation type into multi classification based on Feature Fusion. Extracting knowledge from multi-source heterogeneous data that is beneficial to target tasks, fusion processing at the feature level and identificating the fishing vessel operation based on Feature Fusion. Based on the fishing vessel operation data from Beidou VMS system, it combined with the data of policies and documents of the relevant fishery regulatory authorities, it extracted feature information and constructing feature fusion space for target task to learn model about identification of the fishing vessel operation type. The experimental results on the data of fishing boat provided by Beidou VMS system show that algorithms based on feature fusion has better identification performance than that based on single data source and the identification accuracy of fishing boat operation type is significantly improved after feature fusion. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|