Multi-resolution analysis with frontal decomposition

Use of the compactlyB-spline wavelet of Chui and Wang (1991); Chui (1992) is hindered by loss of accuracy on decomposition, through truncation of weight sequences which are countably infinite. Adaptations to finite intervals often encounter problems at boundaries. For multiresolution analysis on a finite interval employing the linearB-wavelet the present research provides a frontal approach to decomposition which avoids truncation of weight sequences, experiences no problems at boundaries, and which is exhibits a factor of three increase in computational efficiency. The boundary wavelets which complete the linearB-wavelet basis on a finite interval are constructed.