grains is a software package for the tens of tech savy metallurgists and ceramists out there who are tired of counting grains by hand. The goal of this project is to accelerate the process of grain analysis by applying modern computer vision algorithms. This package does not yet have a stable release.
This example showcases the basic functionality of the grains software library. Please note that the scale in this example is ARBITRARY as I have not yet found adequate validation images. Refer to the examples directory for the simple source code used to generate these results.
Centroids Image - Default Settings
Area Histogram - Default Settings
Text Summarization - Default Settings
Automatically Generated by grains Input filename: test_grains.jpg Number of grains: 40 Grain area mean: 1631.55 um^2 Grain area variance: 1527136.86 um^2 Grain area standard deviation: 1235.77 um^2
Grain counter with OpenCV
Size distribution analysis
Calculate average grain size in accordance with ASTM E112
Validate performance against relevant literature
Wrap functionality into a simple CLI
Package for pip install