A selfie is typically a self-portrait captured using the front camera of a smartphone. Most state-of-the-art smartphones are equipped with a high-resolution (HR) rear camera and a low-resolution (LR) front camera. As selfies are captured by front camera with limited pixel resolution, the fine details in it are explicitly missed. This paper aims to improve the resolution of selfies by exploiting the fine details in HR images captured by rear camera using an example-based super-resolution (SR) algorithm. HR images captured by rear camera carry significant fine details and are used as an exemplar to train an optimal matrix-value regression (MVR) operator. The MVR operator serves as an image-pair priori which learns the correspondence between the LR-HR patch-pairs and is effectively used to super-resolve LR selfie images. The proposed MVR algorithm avoids vectorization of image patch-pairs and preserves image-level information during both learning and recovering process. The proposed algorithm is evaluated for its efficiency and effectiveness both qualitatively and quantitatively with other state-of-the-art SR algorithms. The results validate that the proposed algorithm is efficient as it requires less than 3 seconds to super-resolve LR selfie and is effective as it preserves sharp details without introducing any counterfeit fine details.

No abstract

No abstract

In this paper, we propose an efficient noise reduction method that can be used to reduce speckle and jointly enhancing the edge information, rather than just inhibiting smoothing. In this method speckle is removed by filtering of band pass ultrasound images in Laplacian pyramid domain by using mixed PDE based nonlinear diffusion. In each pyramid layer, a gradient threshold is estimated automatically using robust median estimator. The mean absolute error (MAE) between two adjacent diffusion steps is used as stopping criterion. Quantitative results on synthetic data and simulated phantom show the performance of the proposed method compared to state of the art methods. Results on real images demonstrate that the proposed method is able to preserve edges & structural details of the image.

In this work, a compound PDE approach is proposed in multiscale for speckle reduction of diagnostic Ultrasound (US) images, Satellite aperture radar (SAR) images and optical coherence tomography (OCT) images. In denoising process, it is always difficult to preserve discontinuities in one part of the image and simultaneously recovery of smooth areas in other part of image. Hence, combining different algorithms is the only way to improve the image restoration capability. In this paper, coupled PDE, complex diffusion and second order non linear diffusion are applied to layers 1, 2 and 3 of Laplacian pyramid respectively. . In each pyramid layer, using robust median estimator, gradient threshold is estimated automatically. To limit the number of iterations, mean absolute error (MAE) between two adjacent diffusion steps is used as stopping criteria. Quantitative results on synthetic data and simulated phantom show the performance of the proposed method over state of the art methods. Results on real images demonstrate that the proposed method effectively suppresses the speckle and preserves edges & structural details of the image.