Amplitude-based dominant component analysis for underwater mines extraction in side scans sonar

For the first time, the application of the amplitude dominant component analysis (ADCA) to the segmentation of sonar images is explored. We exploit the saliency of the objects in side scans sonar images for underwater mines recognition. Due to the textural and multicomponent nature of the sonar image, a set of bandpass filters is used to decompose the image into narrowband components which lends itself more easily to analysis. The filters bank used is a set of Gabor filters, favored due to their optimal joint spatial and spectral localization. The ADCA-based segmentation is illustrated on real high-resolution sonar images, yielding very promising results showing the interest to exploit the saliency of sonar images.

[1]  J. Daugman.  Uncertainty relation for resolution in space, spatial frequency, and orientation optimized by two-dimensional visual cortical filters. , 1985, Journal of the Optical Society of America. A, Optics and image science.

[2]  Anil K. Jain,et al.  Unsupervised texture segmentation using Gabor filters , 1990, 1990 IEEE International Conference on Systems, Man, and Cybernetics Conference Proceedings.

[3]  Alan C. Bovik,et al.  Multidimensional quasi-eigenfunction approximations and multicomponent AM-FM models , 2000, IEEE Trans. Image Process..

[4]  Marios S. Pattichis,et al.  Active contour segmentation guided by AM-FM dominant component analysis , 2001, Proceedings 2001 International Conference on Image Processing (Cat. No.01CH37205).

[5]  John K. Tsotsos,et al.  Saliency Based on Information Maximization , 2005, NIPS.

[6]  Petros Maragos,et al.  Dominant spatio-temporal modulations and energy tracking in videos: Application to interest point detection for action recognition , 2012, 2012 19th IEEE International Conference on Image Processing.

[7]  Abdel-Ouahab Boudraa,et al.  An unsupervised sonar images segmentation approach , 2007, VISAPP.

[8]  Benoit Zerr,et al.  Sidescan sonar imagery segmentation with a combination of texture and spectral analysis , 2013, 2013 MTS/IEEE OCEANS - Bergen.

[9]  Scott Reed,et al.  An automatic approach to the detection and extraction of mine features in sidescan sonar , 2003 .

[10]  Patrick Pérez,et al.  Sonar image segmentation using an unsupervised hierarchical MRF model , 2000, IEEE Trans. Image Process..

[11]  Christian Knauer,et al.  Target detection of man made objects in side scan sonar images ‐ segmentation based false alarm reduction ‐ , 2008 .

[12]  Peyman Milanfar,et al.  Visual saliency for automatic target detection, boundary detection, and image quality assessment , 2010, 2010 IEEE International Conference on Acoustics, Speech and Signal Processing.

[13]  Christof Koch,et al.  A Model of Saliency-Based Visual Attention for Rapid Scene Analysis , 2009 .