Edge-preserving block compressive sensing with projected landweber

Abstract

Compressive Sensing (CS) is an emerging new sampling technique which helps to break through the Nyquist sampling frequency for sparse signals. This paper addresses improving one of its recovery algorithms known as the Block Compressive Sensing with Smooth Projected Landweber (BCS-SPL). For reducing the blocking artifacts in BCS-SPL, the Wiener filter has been implemented as a classic way to smooth image at the beginning of each iteration, but it is quite sensitive to image edges and blurs the image. In this paper, we propose a modified method which separates image signal into its low and high frequency components, and then independently processes each of the two components. Subsequently, a smoothness enhancing operation is implemented to improve reduction of high frequency oscillatory artifacts after hard thresholding. Experimental results show that the proposed method improves reconstructed image quality by more than 3dB compared to the conventional BCS-SPL.

Automatic number plate recognition on electronic toll collection systems for Vietnamese conditions

“…Among many low-pass filters (e.g. Wiener filter, Median filter, Average filter, Gaussian filter), the Gaussian filter is a good candidate because of its good performance and easiness in controlling the degree of smoothness via the standard deviation [15]. …”

Compressive Sensing Recovery of Natural Images Using Smooth Residual Error Regularization

“…In CS recovery, the reconstructed image desires to be close to the original image, so some related prior information is preferred to be used [5][6][7][8][9][10][11][12]. The pseudo inversed measurement error of   , that is,…”

“…-Based on the Dantzig selector, we propose a new regularization term using a smooth filter like Gaussian filter [8,14,19] or nonlocal means (NLM) filter [18] , called smooth residual error (SRE) regularization. This regularization not only suppresses noise and artifacts in reconstructed images, but also makes the reconstructed image closer to the original image.…”

“…Gaussian filter is a classical filter typically used for reducing noise and artifacts [8,12,19] . The Gaussian filter is widely used due to its simplicity and effectiveness for images suffering much from noise.…”

“…In fact, it is based on the assumption of piecewise smoothness [8,12,19] . In a natural image, this type of assumption is quite valid in smooth regions, but not quite so in a non-stationary regions near edge objects.…”