A Fast Hierarchical MRF sonar Image Segmentation Algorithm

Hierarchical Markov random field (MRF) algorithm has been widely used in sonar image segmentation. To improve the algorithm speed, we abandon the method which describes different regions using regional distribution of grey distribution model and propose a new method in which sonar image was described using grey value statistics. In the proposed method, considering the characteristic of sonar image and the flexibility of this method, the number of category for segmentation can be changed from three to two to improve the calculation speed. They are region of interest (ROI) and background regions respectively. Firstly, C-Means algorithm is used to pre-segment image into three categories. The target and shadow regions are classified as ROI. Secondly, hierarchical MRF model is deployed to segment images into two regions. Finally, the ROI is re-segmented into two regions using greyscale threshold method. Results show that the proposed method can effectively reduce model parameters and computation cost, and the real-time segmentation of sonar image can be achieved.

[1]  Jong Hwan Lim,et al.  Effective Localization of Mobile Robots Using a Sonar Sensor Ring , 2010, Int. J. Robotics Autom..

[2]  Ralph Etienne-Cummings,et al.  A Time-Series Novelty Detection Chip for Sonar , 2004, Int. J. Robotics Autom..

[3]  Bo Du,et al.  Random-Selection-Based Anomaly Detector for Hyperspectral Imagery , 2011, IEEE Transactions on Geoscience and Remote Sensing.

[4]  Patrick Pérez,et al.  Discrete Markov image modeling and inference on the quadtree , 2000, IEEE Trans. Image Process..

[5]  Hong Yu Bian,et al.  Sonar Image Segmentation Using the Level Set Method without Re-Initialization , 2014 .

[6]  D. M. Titterington,et al.  t -Tests, F -Tests and Otsu's Methods for Image Thresholding , 2011, IEEE Trans. Image Process..

[7]  Max Mignotte,et al.  Endocardial Boundary E timation and Tracking in Echocardiographic Images using Deformable Template and Markov Random Fields , 2001, Pattern Analysis & Applications.

[8]  Syh-Shiuh Yeh,et al.  Target Position estimation using Multi-Vision System Implemented on Distributed Mobile robots , 2013, Int. J. Robotics Autom..

[9]  Peter X. Liu,et al.  Sonar image segmentation based on GMRF and level-set models , 2010 .

[10]  M. Lianantonakis,et al.  Sidescan Sonar Segmentation Using Texture Descriptors and Active Contours , 2007, IEEE Journal of Oceanic Engineering.

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

[12]  Giuseppe Oriolo,et al.  An Adaptive Scheme for Image-Based Visual servoing of an underactuated UAV , 2014, Int. J. Robotics Autom..

[13]  Son-Cheol Yu,et al.  Real-time sonar image enhancement for AUV-based acoustic vision , 2015 .

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

[15]  Enrique Coiras,et al.  Automatic Target Recognition in Synthetic Aperture Sonar Images Based on Geometrical Feature Extraction , 2009, EURASIP J. Adv. Signal Process..

[16]  Bo Du,et al.  Spatial Coherence-Based Batch-Mode Active Learning for Remote Sensing Image Classification , 2015, IEEE Transactions on Image Processing.

[17]  Christophe Collet,et al.  Hierarchical MRF modeling for sonar picture segmentation , 1996, Proceedings of 3rd IEEE International Conference on Image Processing.