| 摘要 |
In a method of image segmentation, a tree is derived from input data (310). A no-fill ordering of a matrix is produced, the matrix representing the tree (320). A system of linear equations is solved with respect to the matrix, obtaining a solution (330). The solution is used to define a family of segmentations (340). A segmentation is selected from the family of segmentations based on a metric of segment quality (350). |
