Biomedical subjects
Publications and source records attributed to David Casasent.
The image processing used in the automated analysis of grain boundaries and triple junctions in scanning electron microscopy images is described. The required image processing includes the location of grain boundaries and triple junctions, calculation of the dihedral angles at triple junctions, and selection of electron backscatter probe points (to obtain grain orientation data).
We propose a novel technique for the design of radial basis function (RBF) neural networks (NNs). To select various RBF parameters, the class membership information of training samples is utilized to produce new cluster classes. This allows emphasis of classification performance for certain class data rather than best overall classification. This allows us to control performance as desired and to approximate Neyman-Pearson classification. We also show that by properly choosing the desired output neuron levels, then the RBF hidden to output layer performs Fisher discrimination analysis, and that the full system performs a nonlinear Fisher analysis. Data on an agricultural product inspection problem and on synthetic data confirm the effectiveness of these methods.