Self-adaptive seismic data reconstruction and denoising using dictionary learning based on morphological component analysis

Data reconstruction and data denoising are two critical preliminary steps in seismic processing. Compressed Sensing states that a signal can be recovered by series of solving algorithms if it is sparse transform domain, has been well applied the field reconstruction, when, representation key point. Considering complexity diversity data, single mathematical transformation will lead to incomplete expression bad restoration effects. Morphological Component Analysis (MCA) decomposes into several components with outstanding morphological features approximate complex internal structure. However, ability combined dictionaries constrained number dictionaries, cannot self-adaptively matched features. Dictionary learning overcomes limitation fixed base function training fully suitable for processed but requires huge amount time considerable hardware cost. To solve above problems, new dictionary library (K-Singluar Value Decomposition Discrete Cosine Transform dictionary) hereby proposed based on efficiency high precision dictionary. The self-adaptive achieved under framework successfully data. Real tests have proved method performs better than other dictionaries.