Defra - Department for Environment, Food and Rural Affairs.

Science Search

Science and Research Projects

Return to Science Search homepage   Return to Project List

Use of Machine Vision for Assessment of Cereal Germination Phase IIA - VS04002

Description
Abstract of Research
This research programme is based on the furtherance of an existing DEFRA funded strategic development project, 'The use of machine vision for automated assessment and reporting of cereal seed germination'. This Phase I project is due to complete at the beginning of July 2002. The results to date are summarised in an annual summary sent to DEFRA in July 2001 and confirm that the approach utilised has merits and great potential in a wide class of problems as well as the specific one of seed germination studied. The current research has so far: identified a cost-effective method of sample presentation and image capture; optimised background removal through the use of intelligent algorithms; utilised fuzzy classification systems to optimise seedling segmentation into component parts of roots and seeds; and produced a demonstration CD-ROM of the image analysis tools and algorithms identified so far. The remainder of the project will complete the optimisation of the seedling segmentation process to include shoots and improve the speed and rigour of the process.
This application for funding under DEFRA strategic development seeks to extend the ‘proof of concept’ a further step to the ‘decision rule’ phase (Phase IIA and Phase IIB). Thus the research programme detailed here is the next step along a planned development path as follows:
Phase IIA comprises the initial component of the ‘decision taking part’ of the overall project and will concentrate on the assessment of each of the two outlined approaches.
Phase IIB will extend the assessment of approaches to include artificial neural networks (ANNs) and will take the promising approaches forward for further development and implementation/comparison in a fully working system. Funding for this phase will be applied for to follow on from Phase IIB.
Phase III is a final component comprising technology transfer to develop an efficient ‘automated’ production system. Funding for this phase will be applied for at the end of the Phase II (A&B) project.

Rationale.
Under current statutory seed testing procedures seedling evaluation is carried out according to the ISTA Rules (ISTA, 1999). The interpretation of the germination test involves a process of classification of seeds and seedlings against ISTA defined criteria into dead or dormant seeds and normally or abnormally germinated seedlings. This is a complex process of human interpretation, which requires seed analysts to be highly trained and experienced to ensure reliability of results. This phase II project aims to investigate (Phase IIA) and apply (Phase IIB) machine-based systems to captured and segmented seedling images to replace this human cognitive process.

Project Documents
• Executive Summary : Use of Machine Vision for Assessment of Cereal Germination Phase IIA   (1424k)
Time-Scale and Cost
From: 2003

To: 2004

Cost: £10,000
Contractor / Funded Organisations
National Institute of Agricultural Botany
Keywords
Cereals              
Certification              
Plant Varieties and Seeds              
Plants and Animals              
SeedTesting              
Fields of Study
Plant Varieties and Seeds