An active contour computer algorithm for the classification of cucumbers

The cucumber is one of the most important crops worldwide and, because it is generally consumed fresh, it must be classified into quality categories. The European classification system includes a parameter that relates the degree of curvature relative to the length. Until now, this classification co...

Full description

Saved in:
Bibliographic Details
Published in:Computers and electronics in agriculture Vol. 92; pp. 75 - 81
Main Authors: Clement, Javier, Novas, Nuria, Gazquez, José-Antonio, Manzano-Agugliaro, Francisco
Format: Journal Article
Language:English
Published: Amsterdam Elsevier B.V 01.03.2013
Elsevier
Subjects:
Agronomy. Soil science and plant productions
algorithms
Artificial vision
Biological and medical sciences
crops
Cucumber
cucumbers
Curvature
Fundamental and applied biological sciences. Psychology
Grading
humans
Length
Shape
Shape
Grading
Length
Cucumber
Curvature
Artificial vision
Cucurbitaceae
Cucumis sativus
Algorithm
Vegetable crop
Dicotyledones
Angiospermae
Classification
Spermatophyta
Active contour
ISSN:0168-1699, 1872-7107
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:The cucumber is one of the most important crops worldwide and, because it is generally consumed fresh, it must be classified into quality categories. The European classification system includes a parameter that relates the degree of curvature relative to the length. Until now, this classification could not been be achieved with an automatic system due to the difficulty associated with correctly calculating the axis of a cucumber. This article describes a computer algorithm that uses active contours or “snakes” to classify cucumbers by length and curvature. This algorithm demonstrates an advantage in the determination of the central line of each cucumber, based on an iterative process that is quick and carries out the classification process efficiently. The method was validated against human classification for 360cucumbers and was also compared with an ellipsoid approximation method. The active contour method reduced the classification error by 15% points, compared with the ellipsoid approximation method, to 1%, with no serious errors (i.e., misclassification of Class Extra and I into Class II or vice versa). Meanwhile, the ellipsoid approximation method led to a 16% error rate, of which 2% were serious errors (an error of two classes). The developed approach is applicable to fresh cucumber commercial classification lines to meet the requirements of the European regulations for cucumber classification.
Bibliography:http://dx.doi.org/10.1016/j.compag.2013.01.006
ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 23
ISSN:0168-1699
1872-7107
DOI:10.1016/j.compag.2013.01.006