On-line identification of seeds in mandarins with magnetic resonance imaging

Hernández Sánchez, Natalia and Barreiro Elorza, Pilar and Ruiz Cabello Osuna, Jesus Maria (2006). On-line identification of seeds in mandarins with magnetic resonance imaging. "Biosystems Engineering", v. 95 (n. 4); pp. 529-536. ISSN 1537-5110. https://doi.org/10.1016/j.biosystemseng.2006.08.011.

Description

Title: On-line identification of seeds in mandarins with magnetic resonance imaging
Author/s:
  • Hernández Sánchez, Natalia
  • Barreiro Elorza, Pilar
  • Ruiz Cabello Osuna, Jesus Maria
Item Type: Article
Título de Revista/Publicación: Biosystems Engineering
Date: December 2006
ISSN: 1537-5110
Volume: 95
Subjects:
Faculty: E.T.S.I. Agrónomos (UPM) [antigua denominación]
Department: Ingeniería Rural [hasta 2014]
UPM's Research Group: LPF-TAGRALIA
Creative Commons Licenses: Recognition - No derivative works - Non commercial

Full text

[img]
Preview
PDF - Requires a PDF viewer, such as GSview, Xpdf or Adobe Acrobat Reader
Download (452kB) | Preview

Abstract

Mandarins have been inspected using magnetic resonance imaging (MRI) in order to detect the presence of seeds. To enhance contrast between seeds and pulp, effective transverse relaxation time-weighted fast low angle shot images (703 ms acquisition time) were acquired. Stationary fruits were imaged and then the images were segmented to extract several features. The maximum radius of the region containing the seeds and the central axis rmax, and the perimeter of this region P were the most powerful features for discrimination between seedless and seed-containing fruits. Such features were the most robust since they showed the lowest noise-to-signal ratios (N/S). The proportions of correct classification were 88.9% and 86.7% for seedless and seed-containing fruits, respectively, under MRI stationary conditions. The performance under on-line conditions was evaluated by imaging the fruits while conveyed at 54 mm/s. An analysis of variance with the features extracted from the static images and the motion-corrected dynamic images showed that there were statistically indistinguishable. The proportions of correct classification were 92.5% and 79.5% for the seedless and seed-containing category, respectively, under MRI dynamic conditions. Reduction in the distance between categories for rmax was addressed as the main cause for the decrease in discrimination performance. The robustness of the motion correction procedure was highlighted by the low differences in the N/S ratio and the noise-to-measured range ratios between static and dynamic features.

More information

Item ID: 5016
DC Identifier: http://oa.upm.es/5016/
OAI Identifier: oai:oa.upm.es:5016
DOI: 10.1016/j.biosystemseng.2006.08.011
Official URL: http://www.sciencedirect.com/science/journal/15375110
Deposited by: Memoria Investigacion
Deposited on: 18 Nov 2010 11:18
Last Modified: 20 Apr 2016 14:00
  • Logo InvestigaM (UPM)
  • Logo GEOUP4
  • Logo Open Access
  • Open Access
  • Logo Sherpa/Romeo
    Check whether the anglo-saxon journal in which you have published an article allows you to also publish it under open access.
  • Logo Dulcinea
    Check whether the spanish journal in which you have published an article allows you to also publish it under open access.
  • Logo de Recolecta
  • Logo del Observatorio I+D+i UPM
  • Logo de OpenCourseWare UPM