In this paper a new morphological spectrum image analysis approach using mathematical morphology operation is presented. This approach represents complex shapes by its image spectrum components. The principal advantages of this approach are shift-invariant

high data compression rate

easy real-time implementation

using for invoking the shape filtering and shape measuring etc.

the original binary image can be reconstructed exactly from its morphological shape spectrum.