USGS · 70156415
An automated and universal method for measuring mean grain size from a digital image of sediment
Abstract
Existing methods for estimating mean grain size of sediment in an image require either complicated sequences of image processing (filtering, edge detection, segmentation, etc.) or statistical procedures involving calibration. We present a new approach which uses Fourier methods to calculate grain size directly from the image without requiring calibration. Based on analysis of over 450 images, we found the accuracy to be within approximately 16% across the full range from silt to pebbles. Accuracy is comparable to, or better than, existing digital methods. The new method, in conjunction with recent advances in technology for taking appropriate images of sediment in a range of natural environments, promises to revolutionize the logistics and speed at which grain-size data may be obtained from the field.
Explore related subjects
Keep this discovery
Explore connections, maps & timelines
Daniel D. Buscombe, David M. Rubin, Jonathan A. Warrick. 2010. An automated and universal method for measuring mean grain size from a digital image of sediment. https://pubs.usgs.gov/publication/70156415
Cite the original work for its findings. Save a collection to share your selection of sources.