Suppose I use laplacian pyramid for image denoising application, how would it be better than wavelets? I have read some documents related to laplacian tools in which laplacian pyramids are said to have better selection for signal decomposition than wavelets.
I do not exactly know how laplacian pyramid works, but there are other methods that have demostrated being better than wavelet for multiresolution decomposition, e.g., the Hermite Transform. The problem in using wavelets is that they are not good to represent images curves because the decomposition is made only in three orientations (horizontal, vertical and diagonal).
Each method as it own advantage and disadvantage. I think laplacian pyramid will be better it as advantage of getting line information. At the same time decomposition factor and decimation need to work out properly.
- Badges
- Science topic
- Similar topics
- Engineering
- Electrical Engineering
- Signal Processing
- Image Processing