In the code provided above, the Haar wavelet low-pass filter is already computed as ‘Scale_filter’. We also need to compute the high-pass filter to eventually compute the LH, HL and the HH sub-bands. Please refer to the code snippet below as a reference to generate a Haar wavelet high-pass filter-
% High-pass filter
Wavelet_filter = [-1/sqrt(2) 1/sqrt(2)];
After computing the same, we can apply the appropriate convolutions of low-pass and high-pass filters along rows and columns according to the type of filter required. Please refer to a brief overview of the appropriate convolutions below. You can try them in accordance with your code to implement the remaining sub-bands-
- LH: Low-pass filter on columns, high-pass filter on rows.
- HL: High-pass filter on columns, low-pass filter on rows.
- HH: High-pass filter on both rows and columns.
The remaining steps of downsampling and periodic extension in the code provided can be kept the same while implementing these filters.
I hope the above resolve your query.
