Shokubutsu Kojo Gakkaishi
Online ISSN : 1880-3555
Print ISSN : 0918-6638
ISSN-L : 0918-6638
Papers
Measurement of Lettuce Growth by Image Analysis
—Discrimination of the Number of Days After Planting by Linear Discriminant Methods—
Hiromichi ITOH
Author information
JOURNAL FREE ACCESS

2004 Volume 16 Issue 4 Pages 183-192

Details
Abstract

This research was conducted to develop an algorithm to recognize the number of days after planting of lettuce in a plant factory. Speaking Plant Approach (SPA) has been recognized in a growth control system. Growth information should be utilized actively for environmental control. An instrumentation that can acquire the growth information is necessary to incorporate the SPA into the growth control system. This research is a fundamental analysis for a growth monitoring method. Two varieties of lettuces, the Telme lettuce and the Tango lettuce, were examined. Digital images of the lettuces were photographed by a digital still camera. Several image features, such as the top projected leaf area, the color attributes and the texture features, were calculated using an image processing algorithm. Linear discriminant analysis was adopted for the image features. This research focused on three dates in a growth term. As for the Telme lettuce, the 10th, 20th and 30th days after planting were set as the target dates to note. For the Tango lettuce, the 7th, 14th and 21st days were fixed as objectives. The developed algorithm showed a 2.5% error rate for the Telme lettuce and 20% for the Tango lettuce. Therefore, this recognition method could be utilized for a growth diagnosis of the Telme lettuce. On the other hand, for the Tango lettuce, it is necessary for the algorithm to be further improved.

Content from these authors
© 2004 by Japanese Society of High Technology in Agriculture
Previous article Next article
feedback
Top