A neural network based pattern recognition system for somatic embryos of Douglas fir

摘要: A pattern recognition system was developed to classify Douglas fir somatic embryos by employing an image analysis and two neural network based classifiers. The contour of embryo images segmented, digitalized converted numerical values after the discrete fast Fourier transformation. These values, or features, along with some other shape factors, were used for classification. a hierarchical decision tree into three normal one abnormal classes. An accuracy greater than 80% achieved embryos. This provides objective efficient method classifying fir. It will be useful tool kinetic studies process optimization conifer embryogenesis.