Watersheds in Digital Spaces: An Efficient Algorithm Based on Immersion Simulations
IEEE Transactions on Pattern Analysis and Machine Intelligence
Geodesic Saliency of Watershed Contours and Hierarchical Segmentation
IEEE Transactions on Pattern Analysis and Machine Intelligence
Hi-index | 0.00 |
A buried land mine is one of the most difficult problems faced during and after the war. The most serious problem in mine detection application is the ambiguity of target due to low contrast and background clutter. In this work, the mine detection problem is solved in the context of pre-processing and segmentation techniques for the data associated with infrared and infrared polarization sensors. Principle Component Analysis as a dynamic pr-processing is used to extract the whole dynamic information contained in a sequence of images. Also, the paper proposes a new hierarchical segmentation based on watershed for discriminating land mine from background clutter. The results indicate that the watershed is suitable to segment the outdoor images into noticeable texture region and gives good results for mine detection in IR polarization images. On contrary, the new hierarchy watershed encounters some difficulty in distinguishing buried mines from clutter in IR image.