Subspace Design of Compactly Supported Orthonormal Wavelets

We derive a new matrix parameterization of compactly supported orthonormal wavelets where the coefficients of the wavelet filter are the solution of a linear system of equations that is parameterized by an arbitrary vector. The parameterization shows that the vector of the wavelet filter coefficients is the kernel of a subspace of the condition matrix row-space. This property is exploited to develop a new design procedure for orthonormal wavelets of compact support. The proposed parameterization also describes the class of two-channel orthogonal filter banks where in this case we have two extra degrees of freedom in the design. The effectiveness of the proposed procedure is illustrated by design examples of common orthonormal wavelets.