An automatic alarm method for buried targets based on membership classification

Because there are many false targets in the seabed imaging of SAS (synthetic aperture sonar), it is difficult for the automatic alarm of buried column targets. An automatic alarm method for buried targets based on membership classification is proposed in this paper. Firstly, the mean-standard deviat...

Full description

Saved in:
Bibliographic Details
Published in2017 IEEE International Conference on Signal Processing, Communications and Computing (ICSPCC) pp. 1 - 4
Main Authors Huanhuan Xue, Weihua Cong, Bibo Zhu
Format Conference Proceeding
LanguageEnglish
Published IEEE 01.10.2017
Subjects
automatic alarm
buried column targets
Entropy
Feature extraction
Image segmentation
mean-standard deviation maximum entropy
membership
SAS image
Shape
Synthetic aperture sonar
Online AccessGet full text

Cover

More Information
Summary:Because there are many false targets in the seabed imaging of SAS (synthetic aperture sonar), it is difficult for the automatic alarm of buried column targets. An automatic alarm method for buried targets based on membership classification is proposed in this paper. Firstly, the mean-standard deviation maximum entropy segmentation is used to segment seabed image. Then the area and posture ratio of the segmented target are extracted, and the column membership is calculated. Finally, the automatic alarm of column target is realized through column membership discrimination. The real data processing results show that the method can effectively realize the automatic alarm of buried column targets in SAS images.
DOI:10.1109/ICSPCC.2017.8242496