Extended Basis Pursuit Model and Its Application in Image De-noising

Traditional Basis Pursuit model is adapted to signal de-noising under additive Gaussian noise. Based on the different fitness error term, one new kind Extended Basis Pursuit De-Noising (EBPDN) model is brought forward and applied into salt-and-pepper noise removal. A comparison study of performance of the median filter, the peak-and-valley filter, the detail preserving filter and the EBPDN model is carried out using different types of images. EBPDN model can provide good de-noising results and outperforms other filters in terms of noise suppression and detail preservation.