FEASIBILITY OF A MACHINE-VISION-BASED GRAIN CLEANER

Abstract

A lab scale grain cleaning system was designed, fabricated, and evaluated for its feasibility to be integrated witha machine vision system. It was hypothesized that the ranges of machine-vision -measured features derived from grain samplesbefore and after passing them through the cleaner can be used to evaluate the cleaning performance. Algorithms werewritten to quantify the amount of impurity present in a grain sample only on the basis of machine-vision -extracted featureswithout physically weighing the samples. This machine-vision -based performance was compared with the physical weighingtechnique and the results were statistically similar.

Identifiers

DOI

10.13031/2013.15880

Citations and references