Multi model LSTM architecture for Track Association based on Automatic Identification System Data

Md Asif Bin Syed,Imtiaz Ahmed
2023-04-04
Abstract:For decades, track association has been a challenging problem in marine surveillance, which involves the identification and association of vessel observations over time. However, the Automatic Identification System (AIS) has provided a new opportunity for researchers to tackle this problem by offering a large database of dynamic and geo-spatial information of marine vessels. With the availability of such large databases, researchers can now develop sophisticated models and algorithms that leverage the increased availability of data to address the track association challenge effectively. Furthermore, with the advent of deep learning, track association can now be approached as a data-intensive problem. In this study, we propose a Long Short-Term Memory (LSTM) based multi-model framework for track association. LSTM is a recurrent neural network architecture that is capable of processing multivariate temporal data collected over time in a sequential manner, enabling it to predict current vessel locations from historical observations. Based on these predictions, a geodesic distance based similarity metric is then utilized to associate the unclassified observations to their true tracks (vessels). We evaluate the performance of our approach using standard performance metrics, such as precision, recall, and F1 score, which provide a comprehensive summary of the accuracy of the proposed framework.
Machine Learning
What problem does this paper attempt to address?
### What problem does this paper attempt to solve? This paper primarily addresses the problem of Track Association in maritime surveillance. Specifically: 1. **Background**: - Track Association refers to the problem of matching vessel observation data with their true trajectories in maritime surveillance. - The Automatic Identification System (AIS) provides a large amount of dynamic and geospatial information, offering researchers a new opportunity to tackle this problem. 2. **Research Motivation**: - With the increasing availability of AIS data, researchers can develop sophisticated models and algorithms to effectively handle the challenges of track association. - The advancement of deep learning technologies allows track association to be treated as a data-intensive problem. 3. **Methodology**: - A multi-model framework based on Long Short-Term Memory (LSTM) networks is proposed for track association. - LSTM is a type of recurrent neural network architecture capable of handling multivariate time-series data collected over time and predicting the current position of vessels. - The prediction results and geocentric distance similarity measures are used to associate unclassified observation data with true trajectories. 4. **Experimental Validation**: - The performance of the proposed method is evaluated using standard performance metrics such as precision, recall, and F1 score. - Experimental results show that the model performs well in most cases, but further improvements are needed for certain specific vessels. In summary, this paper aims to solve the problem of track association based on AIS data using LSTM networks and validates its effectiveness through experiments.