hi

Advantages

Most of the energy resides in low frequency components – A good approximation of the original image can be reconstructed using a few components .

Many of the transform coefficients have low associated energies and can be discarded or coarsely quantized with little image distortion .

Provides optimal compression performance from an energy compaction perspective .

– Hardware acceleration available on CPUs

- Computational simple .

Since all operations are independent, they can be performed in parallel to improve computational efficiency

Disadvantages

Computationally expensive, since the transform is data-dependent and deriving basis functions are non-trivial .

Relatively poor compression performance