FastSegmentation

The FastSegmentation package provides a method for automatic and unsupervised cell image segmentation.

Installation

You can install the development version of FastSegmentation like so:

library(devtools)
devtools::install_github("blytheking/FastSegmentation")

Segmentation Example

Segmentation by fast GP can be performed using the generate_GP_Masks function, with the output including the smoothed predictive mean, the binary matrix after initial, data-driven thresholding, and the final detected cell masks.

## Perform segmentation
library(FastSegmentation)

#Set path to image to segment
file_path <- "path/to/file/img.jpg"

#Run segmentation function
gp_masks_result <- generate_GP_Masks(file_path)

## Visualize results
library(plot3D)
library(magick)

#Original image
img <- image_read(file_path)
ori_img_matrix <- as.numeric(img[[1]])[,,1]
image2D(ori_img_matrix, main = "Original Image")

#Predictive Mean
image2D(gp_masks_result$combined_predmean, main = "Predictive Mean")

#Thresholding by Criterion 1
image2D(gp_masks_result$combined_thresholded1, main = "Binary Matrix")

#Final Cell Masks
image2D(gp_masks_result, main = "Binary Matrix")

Notes

FastSegmentation uses the EBImage package from BiocManager. Since EBImage is not part of CRAN, please first install it using this code:

library(BiocManager)
BiocManager::install("EBImage")
library(EBImage)

Relevant Literature

This package was adapted from the methods described in the following manuscript:

Baracaldo, L., King, B., Yan, H., Lin, Y., Miolane, N., & Gu, M. (2025). Unsupervised cell segmentation by fast Gaussian processes. arXiv preprint arXiv:2505.18902.