PMID- 37247563
OWN - NLM
STAT- MEDLINE
DCOM- 20230615
LR  - 20230615
IS  - 1879-2057 (Electronic)
IS  - 0001-4575 (Linking)
VI  - 189
DP  - 2023 Sep
TI  - A railway intrusion detection method based on decomposition and semi-supervised 
      learning for accident protection.
PG  - 107124
LID - S0001-4575(23)00171-9 [pii]
LID - 10.1016/j.aap.2023.107124 [doi]
AB  - In recent years, video surveillance has become increasingly popular in railway 
      intrusion detection. However, it is still quite challenging to detect the 
      intruded objects efficiently and accurately because: (a) The backgrounds of video 
      frames generated by the fixed cameras are similar and only few intrusive frames 
      are available, resulting in a lack of diversity among video frames, and further 
      leading to over fitting of the detection models during training; (b) The 
      intrusion of small targets or targets far from the location of camera exhibits 
      sparsity relative to the wide monitoring view of the camera, which challenges the 
      detection of such targets in a complex background; (c) The extreme imbalance 
      between non-intrusive frames and intrusive frames, as well as a large number of 
      unlabeled frames, hinder the effective training of the detection model and weaken 
      its capacity of generalization. To tackle the above issue, this article develops 
      an effective intrusion detection method by combining low-rank and sparse 
      decomposition (LRSD) and Semi-supervised Support Vector Domain Description 
      (Semi-SVDD). Firstly, LRSD is used to decompose the monitored video into a 
      background and a foreground. Then, based on the semantic segmentation method, we 
      extract the mask of the track region in the decomposed background, which is used 
      to mask the foreground. Next, by using both the labeled and unlabeled frames of 
      the masked foreground, Semi-SVDD is established for the intrusion detection. 
      Numerical results show that the removal of background interference and the 
      combination of the labeled and unlabeled information help to improve the 
      performance of the proposed method, and thus superior to benchmark methods.
CI  - Copyright (c) 2023 Elsevier Ltd. All rights reserved.
FAU - Li, Bin
AU  - Li B
AD  - School of Traffic and Transportation, Lanzhou Jiaotong University, Lanzhou 
      730070, PR China; CHN Energy Technology & Economics Research Institute, Beijing 
      102211, PR China. Electronic address: 17230022@chnenergy.com.cn.
FAU - Tan, Lei
AU  - Tan L
AD  - State Key Laboratory of Rail Traffic Control and Safety, Beijing Jiaotong 
      University, Beijing 100044, PR China; Beijing Municipal Engineering Research 
      Institute, Beijing 100037, PR China. Electronic address: 17111038@bjtu.edu.cn.
FAU - Wang, Feng
AU  - Wang F
AD  - State Key Laboratory of Rail Traffic Control and Safety, Beijing Jiaotong 
      University, Beijing 100044, PR China; School of Mechanical, Electronic and 
      Control Engineering, Beijing Jiaotong University, Beijing 100044, PR China. 
      Electronic address: feng.wang@bjtu.edu.cn.
FAU - Liu, Linzhong
AU  - Liu L
AD  - School of Traffic and Transportation, Lanzhou Jiaotong University, Lanzhou 
      730070, PR China. Electronic address: liulinzhong@tsinghua.org.cn.
LA  - eng
PT  - Journal Article
DEP - 20230527
PL  - England
TA  - Accid Anal Prev
JT  - Accident; analysis and prevention
JID - 1254476
SB  - IM
MH  - Humans
MH  - *Algorithms
MH  - *Accidents, Traffic/prevention & control
MH  - Supervised Machine Learning
OTO - NOTNLM
OT  - Decomposition
OT  - Railway intrusion detection
OT  - Semi-supervised learning
OT  - Track regions
COIS- Declaration of Competing Interest The authors declare that they have no known 
      competing financial interests or personal relationships that could have appeared 
      to influence the work reported in this paper.
EDAT- 2023/05/30 01:06
MHDA- 2023/06/15 06:42
CRDT- 2023/05/29 18:05
PHST- 2023/03/30 00:00 [received]
PHST- 2023/05/03 00:00 [revised]
PHST- 2023/05/18 00:00 [accepted]
PHST- 2023/06/15 06:42 [medline]
PHST- 2023/05/30 01:06 [pubmed]
PHST- 2023/05/29 18:05 [entrez]
AID - S0001-4575(23)00171-9 [pii]
AID - 10.1016/j.aap.2023.107124 [doi]
PST - ppublish
SO  - Accid Anal Prev. 2023 Sep;189:107124. doi: 10.1016/j.aap.2023.107124. Epub 2023 
      May 27.