Abstract

Segmenteverygrain is a Python module that addresses the need for quantifying grain size in granular materials. It combines the Segment Anything Model (SAM) with a U-Net-style convolutional neural network to detect and measure grains or clasts in diverse image types, ranging from photomicrographs of sand, mineral grains, and thin sections to photographic images of gravel and boulder fields. Segmenteverygrain supports segmentation of large images, georeferencing, and interactive editing of the results.


Citation

Z. Sylvester et al. (2025). Segmenteverygrain: A Python module for segmentation of grains in images. Journal of Open Source Software v. 10.

@article{sylvester_segmenteverygrain_2025,
	title = {Segmenteverygrain: {A} {Python} module for segmentation of grains in images},
	volume = {10},
	issn = {2475-9066},
	shorttitle = {Segmenteverygrain},
	url = {https://joss.theoj.org/papers/10.21105/joss.07953},
	doi = {10.21105/joss.07953},
	language = {en},
	number = {112},
	journal = {Journal of Open Source Software},
	author = {Sylvester, Zoltán and Stockli, Daniel F. and Howes, Nick and Roberts, Kalinda and Malkowski, Matthew A. and Poros, Zsófia and Martindale, Rowan C. and Bai, Wilson},
	month = aug,
	year = {2025}
}