An identification algorithm based on matrix degree of grey incidences

A new classification algorithm based on matrix degree of grey incidences for handwritten number recognition and natural image clustering is proposed in this paper. In order to recognize a handwritten numeral or to classify the type of images, we should extract the features that can describe the differences among all information efficiently ant first. Then the features matrix is build for the samples. By comparing the test samples with the training samples dataset, classification results can be confirmed via the matrix degree of grey incidences between feature matrixes. Experiments show that the proposed algorithm can classify information such as handwritten-numbers and images well.