Liggett Jr, W.
and Fletcher, R.
(2008),
Data Mining Electron Microscope Images to Estimate a Particle Size Distribution, Technometrics
(Accessed April 26, 2024)
Official websites use .gov
A .gov website belongs to an official government organization in the United States.
Secure .gov websites use HTTPS
A lock (
) or https:// means you’ve safely connected to the .gov website. Share sensitive information only on official, secure websites.
Data Mining Electron Microscope Images to Estimate a Particle Size Distribution
Published
Author(s)
,
Abstract
The application considered involves density estimation, spatial prediction, and sizing error. The goal is estimation of a particle size distribution; the data are scanning electron microscope (SEM) images that constitute a sample of a filter surface; and the sizing error is in the measurement of the particles shown in the images. The quantities to be estimated are filter-wide counts of particles in size intervals of interest. Spatial prediction enters because the particle intensity varies smoothly over the filter, and consequently variograms are informative and kriging estimates increase efficiency. Sizing error as a potentially insurmountable problem became apparent during the data analysis. A solution was discovered that depends on the fact that the particle size density has no mode but is monotonically decreasing. Because this problem was not anticipated during project planning, this application became a case study in data mining.Citation
Technometrics
Pub Type
Journals
Keywords
density estimation, measurement error, scanning electron microscope, size fractionation, spatial statistics, spline fitting
Citation
Created October 16, 2008