`skimage.morphology._grayreconstruct:reconstruction_loop`

Signature

   def     reconstruction_loop (    ranks    ,    prev    ,    next    ,    strides    ,    current_idx    ,    image_stride      )

Summary

The inner loop for reconstruction.

Extended Summary

This algorithm uses the rank-order of pixels. If low intensity pixels have a low rank and high intensity pixels have a high rank, then this loop performs reconstruction by dilation. If this ranking is reversed, the result is reconstruction by erosion.

For each pixel in the seed image, check its neighbors. If its neighbor's rank is below that of the current pixel, replace the neighbor's rank with the rank of the current pixel. This dilation is limited by the mask, i.e. the rank at each pixel cannot exceed the mask as that pixel.

Parameters

ranks : array: The rank order of the flattened seed and mask images.
prev, next: arrays: Indices of previous and next pixels in rank sorted order.
strides : array: Strides to neighbors of the current pixel.
current_idx : int: Index of highest-ranked pixel used as starting point in loop.
image_stride : int: Stride between seed image and mask image in aranks.

Aliases

skimage.morphology._grayreconstruct.reconstruction_loop