Discrimination of soybean leaflet shape by neural networks with image input

It is necessary to correctly evaluate intra- and inter-specific variations for the efficient collection and preservation of genetic resources, and lea f shape is one of the important characteristics to be evaluated. It has bee n thought that a mon consistent and quantitative method should be introduce d to aid in the processes of practical discrimination. Several researchers have suggested leaf shape evaluation methods using shape features, and thes e methods have shown good results. The shape features selected in these met hods have differed from one method to another, and new shape features must be redefined when these methods are applied to new cases. The processes for defining and extracting share features are ad hoc. We, therefore, have att empted to develop a generalized model that requires neither the definition nor extraction of any shape features; the method uses neural networks into which leaf shape images are input. In this study, we applied a Hopfield mod el and a simple perceptron to the varietal discrimination of individual lea flet shapes of 364 soybean leaflets of 38 varieties. In the examination of up to ten varieties, the discriminant error of the neural networks with ima ge input was satisfactorily low even under cross validation. We, therefore, concluded that this model works quite well for quantitative varietal discr imination in the case of soybean leaflets. The advantage of requiring neith er the definition nor extraction of any shape features makes us expect that this model will be widely applicable to other cases, and we will attempt t o verify this applicability. (C) 2000 Elsevier Science B.V. All rights rese rved.