Share Email Print

Proceedings Paper

Image processing for x-ray inspection of pistachio nuts
Author(s): David P. Casasent
Format Member Price Non-Member Price
PDF $17.00 $21.00
cover GOOD NEWS! Your organization subscribes to the SPIE Digital Library. You may be able to download this paper for free. Check Access

Paper Abstract

A review is provided of image processing techniques that have been applied to the inspection of pistachio nuts using X-ray images. X-ray sensors provide non-destructive internal product detail not available from other sensors. The primary concern in this data is detecting the presence of worm infestations in nuts, since they have been linked to the presence of aflatoxin. We describe new techniques for segmentation, feature selection, selection of product categories (clusters), classifier design, etc. Specific novel results include: a new segmentation algorithm to produce images of isolated product items; preferable classifier operation (the classifier with the best probability of correct recognition Pc is not best); higher-order discrimination information is present in standard features (thus, high-order features appear useful); classifiers that use new cluster categories of samples achieve improved performance. Results are presented for X-ray images of pistachio nuts; however, all techniques have use in other product inspection applications.

Paper Details

Date Published: 28 March 2001
PDF: 20 pages
Proc. SPIE 10301, Optics in Agriculture: 1990-2000: A Critical Review, 1030109 (28 March 2001); doi: 10.1117/12.420103
Show Author Affiliations
David P. Casasent, Carnegie Mellon Univ. (United States)

Published in SPIE Proceedings Vol. 10301:
Optics in Agriculture: 1990-2000: A Critical Review
James A. DeShazer; George E. Meyer, Editor(s)

© SPIE. Terms of Use
Back to Top