ABSTRACT
It is well known that the discrete Hadamard transform (DHdT) is computationally advantageous over the fast Fourier transform. Being orthonormal and taking value 1 or −1 in each point, the Hadamard functions can be used for a series expansion of the signal. The DHdT has found useful applications in signal and image processing, communication systems, image coding, image enhancement, pattern recognition, and general two-dimensional filtering [4],[65]-[68]. We describe properties of the Hadamard transformation and a class of the so-called bit-and transformations, as well as the class of mixed transformations.
