PlantCellSeg-ImageJ

A modular and practical ImageJ/Fiji pipeline for segmenting plant epidermal cells from 3D confocal microscopy images.
This workflow includes optional preprocessing with SurfCut to better handle curved tissues such as leaves, and uses MorphoLibJ for morphological segmentation.

• Designed for plant epidermal tissue (e.g., leaf surface)
• SurfCut support (optional) for surface flattening
• Morphological segmentation using MorphoLibJ
• ROI extraction for quantitative shape analysis
• ImageJ macro-based, reproducible and extensible

SurfCut extracts the outermost cell layer from a 3D confocal stack by detecting the surface, cropping a user-defined depth, and projecting it into a single 2D image.

📌 Illustration from the SurfCut GitHub repository, used here for educational purposes.

The pipeline processes 3D confocal image stacks of plant tissues with flexible preprocessing options.

• 若葉片彎曲嚴重（如自然捲曲或呈弧形）：建議使用 SurfCut2 Lite
• 若葉片平坦或已壓平：建議使用 Max Projection
• 若影像是單張 2D 或已投影處理過的圖像：可選擇 None

• Applies Despeckle filter and Gaussian blur for noise reduction.

• Uses MorphoLibJ’s automatic watershed segmentation

• Converts segmented label map into vector-based ROIs.

💡 Tip
Upon starting the macro, a dropdown menu will prompt you to choose the preprocessing method, ensuring the pipeline is tailored to your image type.

This project integrates SurfCut2 Lite as a preprocessing step to efficiently and automatically extract the epidermal layer signal from 3D confocal stacks.

• Original SurfCut SurfCut Offers full parameter options and an interactive interface, suitable for in-depth exploration of surface extraction principles and parameter tuning. It is a valuable tool for academic research and method development.

• SurfCut2 Lite
  A lightweight, interface-free version designed for seamless integration into automated ImageJ macro workflows. It greatly saves processing time and improves reproducibility, making it ideal for routine large-scale data analysis.

• If you need flexible parameter adjustments and want to deeply understand the projection mechanism, the original SurfCut is recommended.
• If you prioritize speed, automation, and workflow stability, SurfCut2 Lite is preferred.

• Fiji (ImageJ) with the following plugins installed:
  • MorphoLibJ
  • SurfCut2 Lite — Included in this pipeline
    SurfCut2 Lite is a macro for extracting a surface layer from 3D image stacks,
    especially suited for curved or dome-shaped tissues.
    This macro is integrated into the pipeline and requires no separate download or installation.

The Top and Bottom values in SurfCut2 Lite do not correspond to the original Z-stack slice indices.

Instead, they define the relative depth from the detected surface, after the edge-projection process. Internally, SurfCut generates a surface-aligned mask and shifts it along Z to extract a target layer.

Use integers (in slices), e.g., Top = 10, Bottom = 12, to extract the cell layer approximately 10–12 slices below the curved surface.

📌  These slices are in a synthetic projected space, not physical microns or raw Z indices.

This segmentation result was generated using the Max Projection preprocessing method:

🧩 Segmentation Output (Max Projection)

File: 20160406 3Gp L2 1.tif 📂 Download Example: 20160406 3Gp L2 1.tif

Image file courtesy of Professor Chin-Min Ho, Institute of Plant and Microbial Biology (IPMB).

📌 The Max Projection method is suitable for relatively flat samples or flattened leaf tissue. It may result in blurred boundaries on curved samples.

File: Hypocotyl_GFP-MBD.tif 📂 Download Example: Hypocotyl_GFP-MBD.tif

  

Left: SurfCut projection
Middle: Original max projection
Right: Segmentation result after SurfCut

SurfCut settings used:

• Blur radius: 3
• Threshold: 15
• Top cut: 9
• Bottom cut: 11

Image source (Hypocotyl sample): kindly provided by the authors of SurfCut

• Verger, S., et al. (2019). SurfCut: A macro tool for surface layer extraction from 3D images. BMC Biology, 17, 32.
  https://doi.org/10.1186/s12915-019-0657-1

• Legland, D., Arganda-Carreras, I., & Andrey, P. (2016). MorphoLibJ: integrated library and plugins for mathematical morphology with ImageJ. Bioinformatics, 32(22), 3532–3534.
  https://doi.org/10.1093/bioinformatics/btw413

• Fiji / ImageJ Documentation:
  https://imagej.net/

• SurfCut2 Lite macro:
  https://github.com/VergerLab/SurfCut2

• This pipeline optionally uses SurfCut, which remains under its original license.
• MorphoLibJ is maintained by the INRAE image processing team.
• This pipeline is intended for academic and research use.
• The hypocotyl sample image used in the SurfCut segmentation comparison was generously provided by the authors of the SurfCut project.