Classification of Tissue Culture Segments by Colour Machine Vision

V. Alchanatis, K. Peleg, M. Ziv

Research output: Contribution to journalArticlepeer-review

19 Scopus citations

Abstract

Tissue culture techniques are finding increasingly widespread applications for cloning of many plants. Protocols for mass propagation of many species have been developed, but in spite of its advantages, large-scale commercial plant propagation by tissue cultures is largely limited to ornamental plants. This is due mainly to the intensive skilled labour required for subculturing the propagules and in transferring individual shoots or plantlets into and out of culture containers. In order to cut down the production costs, a certain degree of automation is essential. A cost effective approach for automation is proposed, whereby tissue culture plantlets are chopped into approximately uniformly sized segments, on a conveying production line while using colour computer vision for identifying and locating the number and positions of propagation organs, in images of the plantlet segments. Plantlet segments without propagation organs are rejected, while properly cut segments with viable buds or shoots are automatically selected for subculturing. In this paper, some initial results of this approach are reported, in which stationary images of manually pre-cut potato plantlet segments were analysed and classified. Using colour machine vision and a Neural Network-based classifier, a basis was laid for a practical system, which may be used for automatic classification of tissue culture segments of potato plantlets. Instead of the conventional use of black and white cameras and geometric features, colour features only are used together with colour frame manipulation capabilities, which are now available in most commercial imaging boards. This facilitates accurate, high-speed classification of plantlet images.

Original languageEnglish
Pages (from-to)299-311
Number of pages13
JournalBiosystems Engineering
Volume55
Issue number4
DOIs
StatePublished - Aug 1993

Fingerprint

Dive into the research topics of 'Classification of Tissue Culture Segments by Colour Machine Vision'. Together they form a unique fingerprint.

Cite this