Pattern similarities of vector matrices

Abstract

A concept of vector matrix or matrix of vector is introduced. It is a well-known matrix where the elements are vectors. An example of vector matrix is a digital image represented in a matrix A m×n. The elements of A are vectors in R 3 represent color in red, green, blue, and m × n is the number of pixels. Similarity of two matrices is defined by the similarity of each corresponding elements. Such definition is very strict, two similar matrices must have the same vector for each corresponding elements. Applying for digital images, two similar matrices are actually the same image. Less similarity concept is introduced, namely pattern similarity. Pattern similarity is a generalization of strict similarity by applying a pattern function. An application of pattern similarity is for object recognition.