Two dimensional clutter: a new definition

A great deal of interest has recently been shown in defining clutter for infrared images. Clutter plays an important role in the detection of targets in infrared images. Different research workers have proposed different definitions for clutter. A number of these definitions are based on the variance of cells comprising the image. To the best of author´s knowledge no one has so far proposed a two dimensional definition of clutter. The objective of this paper is to propose a new definition of clutter called clutter matrix. A clutter matrix is a matrix based on the correlation matrix obtained from subcells of an image. The notion of subcells and subcells level is introduced. Under the conventional definitions for clutter, two different images may have same clutter value, but under the proposed definition the distinction between the two images is conspicuous from the two clutter matrices for the images. The central diagonal element of the clutter matrix gives the average of cell variance, i.e. measure for the conventional clutter value for the image. An algorithm for determining clutter matrix is proposed in this paper. The procedure is illustrated with the help of examples made of synthetic and real images