An image-coding scheme which combines transform coding with a human visual system (HVS) model is described. The system includes an eye tracker to pick up the point of regard of a single viewer. One can then utilize that the acuity of the HVS is less in the peripheral vision than in the central part of the visual field. A model of the decreasing acuity of the HVS, which can be applied to a wide class of transform coders is described. Such a coding system has large potential for data compression. In this paper, we have incorporated the model into four different transform coders, one from each of the main classes of transform coders. Two of the coders are block-based decomposition schemes, the discrete cosine transform-based JPEG coder and a lapped transform scheme. The two others are subband-based decomposition schemes, a wavelet based and a wavelet packet-based scheme. © 2002 Elsevier Science B.V. All rights reserved.
Although block compressive sensing (BCS) makes it tractable to sense large-sized images and video, its recovery performance has yet to be significantly improved because its recovered images or video usually suffer from blurred edges, loss of details, and high-frequency oscillatory artifacts, especially at a low subrate. This paper addresses these problems by designing a modified total variation technique that employs multi-block gradient processing, a denoised Lagrangian multiplier, and patch-based sparse representation. In the case of video, the proposed recovery method is able to exploit both spatial and temporal similarities. Simulation results confirm the improved performance of the proposed method for compressive sensing of images and video in terms of both objective and subjective qualities.
The paper discusses which properties of filter sets used in local structure estimation that are the most important. Answers are provided via the introduction of a number of fundamental invariances. Mathematical formulations corresponding to the required invariances leads up to the introduction of a new class of filter sets termed loglets. Loglets are polar separable and have excellent uncertainty properties. The directional part uses a spherical harmonics basis. Using loglets it is shown how the concepts of quadrature and phase can be defined in n-dimensions. It is also shown how a reliable measure of the certainty of the estimate can be obtained by finding the deviation from the signal model manifold. Local structure analysis algorithms are quite complex and involve a lot more than the filters used. This makes comparisons difficult to interpret from a filter point of view. To reduce the number 'free' parameters and target the filter design aspects a number of simple 2D experiments have been carried out. The evaluation supports the claim that loglets are preferable to other designs. In particular it is demonstrated that the loglet approach outperforms a Gaussian derivative approach in resolution and robustness. (c) 2005 Elsevier B.V. All rights reserved.
In this paper, we combine 3D anisotropic diffusion and motion estimation for image denoising and improvement of motion estimation. We compare different continuous isotropic nonlinear and anisotropic diffusion processes, which can be found in literature, with a process especially designed for image sequence denoising for motion estimation. All of these processes initially improve motion estimation due to reduction of noise and high frequencies. But while all the well known processes rapidly destroy or hallucinate motion information, the process brought forward here shows considerably less information loss or violation even at motion boundaries. We show the superior behavior of this process. Further we compare the performance of a standard finite difference diffusion scheme with several schemes using derivative filters optimized for rotation invariance. Using the discrete scheme with least smoothing artifacts we demonstrate the denoising capabilities of this approach. We exploit the motion estimation to derive an automatic stopping criterion.