Seg2Link is a software designed to semi-automatically segment cells in 3D image stacks, especially for the brain images obtained by electron microscopy.

Read the documentation at Seg2Link.io.

Features

• Utilize deep learning predictions -- Seg2Link assist users to create to accurate segmentation results of individual cells from inaccurate cell/non-cell predictions .
• Simplicity -- Seg2Link generates segmentation automatically and allows for easy inspection and manual corrections.
• Efficient -- Seg2Link is designed for the fast processing of medium-sized 3D images with billions of voxels.

Workflow

  flowchart TB
  A[Raw Image]-->|DNN prediction|B[Cell/NonCell]

  subgraph ide1 [Round #1: Seg2D+Link]
    B-->|Segment+Link|C[Linked 2D SEG. in layer i]
    C-->|Manual correction + <br> Seg+Link next layer|C
    C-->D[Linked 2D SEG. in All layers]
  end

  subgraph ide2 [Round #2: 3D correction]
    D-->|Export/Import|E[3D SEG.]
    E-->|Manual <br/> correction|F[Completed SEG]
    F-->|Export|G[TIFF image sequence]
  end
  A-->|Other software|E

Install

• During preparation