Hyperspectral image segmentation using binary partition trees

FREE-DOWNLOAD S Valero, P Salembier… – Submitted to ICIP, 2011

The work presented here proposes a new Binary Partition Tree pruning strategy aimed at the segmentation of hyperspectral images. The BPT is a region-based representation of images that involves a reduced the number of elementary primitive and and therefore allows us to define robust and efficient segmentation algorithm. Here, the regions contained in the BPT branches are studied by recursive spectral graph partitioning. The goal is to remove subtrees composed of nodes which are considered to be similar. To this end, affinity matrices on the tree branches are computed using a new distance-based measure depending on canonical correlations relating principal coordinates. Experimental results have demonstrate the good performances of BPT construction and pruning.