합성곱 오토인코더를 이용한 이상거동 선박 식별
Detection of Abnormal Vessel Trajectories with Convolutional Autoencoder
0
조회
0
다운로드
2
인용
0
참고문헌
키워드
초록
Recently there was an incident that military radars, coastal CCTVs and other surveillance equipment captured a small rubber boat smuggling a group of illegal immigrants into South Korea, but guards on duty failed to notice it until after they reached the shore and fled. After that, the detection of such vessels before it reach to the Korean shore has emerged as an important issue to be solved. In the fields of marine navigation, Automatic Identification System (AIS) is widely equipped in vessels, and the vessels incessantly transmits its position information. In this paper, we propose a method of automatically identifying abnormally behaving vessels with AIS using convolutional autoencoder (CAE). Vessel anomaly detection can be referred to as the process of detecting its trajectory that significantly deviated from the majority of the trajectories. In this method, the normal vessel trajectory is gridded as an image, and CAE are trained with images from historical normal vessel trajectories to reconstruct the input image. Features of normal trajectories are captured into weights in CAE. As a result, images of the trajectories of abnormal behaving vessels are poorly reconstructed and end up with large reconstruction errors. We show how correctly the model detects simulated abnormal trajectories shifted a few pixel from normal trajectories. Since the proposed model identifies abnormally behaving ships using actual AIS data, it is expected to contribute to the strengthening of security level when it is applied to various maritime surveillance systems.
Abstract
Recently there was an incident that military radars, coastal CCTVs and other surveillance equipment captured a small rubber boat smuggling a group of illegal immigrants into South Korea, but guards on duty failed to notice it until after they reached the shore and fled. After that, the detection of such vessels before it reach to the Korean shore has emerged as an important issue to be solved. In the fields of marine navigation, Automatic Identification System (AIS) is widely equipped in vessels, and the vessels incessantly transmits its position information. In this paper, we propose a method of automatically identifying abnormally behaving vessels with AIS using convolutional autoencoder (CAE). Vessel anomaly detection can be referred to as the process of detecting its trajectory that significantly deviated from the majority of the trajectories. In this method, the normal vessel trajectory is gridded as an image, and CAE are trained with images from historical normal vessel trajectories to reconstruct the input image. Features of normal trajectories are captured into weights in CAE. As a result, images of the trajectories of abnormal behaving vessels are poorly reconstructed and end up with large reconstruction errors. We show how correctly the model detects simulated abnormal trajectories shifted a few pixel from normal trajectories. Since the proposed model identifies abnormally behaving ships using actual AIS data, it is expected to contribute to the strengthening of security level when it is applied to various maritime surveillance systems.
참고문헌 (18)
- Besse, P., "Review and Perspective for Distance Clustering of Vehicle Trajectories", IEEE Transactions on Intelligence Transportation Systems, 17(11), pp.3306~3317 (2016)
- Chen, L., "On the Marriage of Lp-norms and Edit Distance", Proceeding of the International Conference on Very Large Data Bases, pp.1040~1049 (2004)
- Chen, L., "Robust and Fast Similarity Search for Moving Object Trajectories", Pro- ceedings of the International Conference on Manage- ment of Data, pp.491~502 (2005)
- Fu, P., "Finding Abnormal Vessel Trajectories Using Feature Learning", IEEE Access, 5, pp.7898~7909 (2017)
- Han, H., "DBSCAN Opti- mization for Improving Marine Trajectory Clustering and Anomaly Detection", International Archives of Photo- grammetry, Remote Sensing and Spatial Information Science, XLIII-B4-2020, pp.445~461 (2020)
- Iltanan, H., "Maritime Anomaly Detection using Autoen- coders and OPTICS-OF", Master, pp.2020
- 김화중, "선박속력 및 급유결정 문제에 대한 휴리스틱 알고리즘", 한국산업경영시스템학회지, 39(2), pp.19~27 (2016), 한국산업경영시스템학회
- 김화중, "Determining Economic Ship Speeds and Fleet Sizes Considering Greenhouse Gas Emissions", 한국산업경영시스템학회지, 34(2), pp.49~59 (2011), 한국산업경영시스템학회
- 김갑기, "AIS를 이용한 선박 안전운항 경보 시스템 개발", 한국정보기술학회논문지, 11(4), pp.1~7 (2013), 한국정보기술학회
- 권세혁, "선박운항 시뮬레이터 실험조건 축소화 연구", 한국산업경영시스템학회지, 35(1), pp.101~106 (2012), 한국산업경영시스템학회
- Nguyen, D., "A Multi-task Deep Learning Architecture for Maritime Surveillance using AIS Data Streams", IEEE 5th International Conference on Data Science and Ad-vanced Analytics(DSAA) (2018)
- 오재용, "AIS 데이터 분석을 통한 이상 거동 선박의 식별에 관한 연구", 한국항해항만학회지, 42(4), pp.277~282 (2018), 한국항해항만학회[DOI]
- 오재용, "오토인코더 모델을 이용한 선박교통관제 지원 시스템의 개발", 정보과학회 컴퓨팅의 실제 논문지, 24(12), pp.642~648 (2018), 한국정보과학회[DOI]
- Palliotta, G., "Data-driven Detection and Context-based Classification of Maritime Anomalies", The 18th International Conference on Information Fusion, pp.1152~1159 (2015)
- Pallotta, G., "Traffic Knowledge Discovery from AIS data", Proceedings of the 16th International Conference on Information Fusion, pp.1996~2003 (2013)
- 박재민, "선박 이상 탐지를 위한 VAE-CUSUM 기반 모니터링 방법론", 대한산업공학회지, 46(4), pp.432~442 (2020), 대한산업공학회[DOI]
- Vlachos, M., "Discovering Similar Multidimensional Trajectories", Proceeding of the International Conference on Data Engineering, pp.673~684 (2002)
- Yu, J.Y., "Deep Learning Approaches for AIS Data Association in the Context of Maritime Domain Awareness", IEEE 23rd International Conference on Information Fusion, pp.1~8 (2020)