openalea.phenomenal.tracking.leaf_extension#

Use binary images to extend the length of each phenomenal phm_leaf.

Method : A 2d skeleton is computed for the binary image at a given angle. Then, the algorithm searches correspondences between phenomenal polylines (reprojected in 2D) and skeleton polylines. For each match, an extension factor e >= 1 is computed this way :

e = (skeleton 2D polyline length) / (phenomenal 2D polyline length).

This is done for each side angle. Then, results are merged : for each phenomenal phm_leaf, the final extension factor is equal to the median of all extension values found for this phm_leaf, or 1 if no extension value was found.

Functions

compute_extension(polylines_phm, polylines_sk)

leaf_extension(phm_seg, binaries, projections)

skeleton_branches(image[, n_kernel, min_length])