Abstract not found

Edges provide semantically important image features. In this paper a lossy compression method for cartoon-like images is presented, which is based on edge information. Edges together with some adjacent grey/colour values are extracted and encoded using a classical edge detector, binary compression standards such as JBIG and stateof-the-art encoders such as PAQ. When decoding, information outside the seen coded data is recovered by solving the Laplace equation, i.e. we inpaint with the steady state of a homogeneous diffusion process. For the discrete reconstruction problem, we prove existence and uniqueness and establish a maximum-minimum principle. Furthermore, we describe an efficient multigrid algorithm. The result is a simple codec that is able to encode and decode in real time. We show that for cartoon-like images this codec can outperform the JPEG standard and

Point trajectories have emerged as a powerful means to obtain high quality and fully unsupervised segmentation of objects in video shots. They can exploit the long term motion difference between objects, but they tend to be sparse due to computational reasons and the difficulty in estimating motion in homogeneous areas. In this paper we introduce a variational method to obtain dense segmentations from such sparse trajectory clusters. Information is propagated with a hierarchical, nonlinear diffusion process that runs in the continuous domain but takes superpixels into account. We show that this process raises the density from 3% to 100 % and even increases the average precision of labels. 1.

frame n 4 frame n-0.5 frame n-1 Figure 1: Remote rendering allows navigating in complex scenes even on weak client hardware. But not only final images are of interest on the client side, auxiliary information like depth or motion become increasingly attractive in this context for various purposes. Examples include spatio-temporal upsampling (1, 2), 3D stereo rendering (3), or frame extrapolation (4). Standard encoders (H.264 in image 1) are currently not always well-adapted to such streams and our contribution is a novel method to efficiently encode and decode augmented video streams with high-quality (compare insets in image 1 and 2). In this paper, we focus on efficient compression and streaming of frames rendered from a dynamic 3D model. Remote rendering and on-the-fly streaming become increasingly attractive for interactive applications. Data is kept confidential and only images are sent to the client. Even if the client’s hardware resources are modest, the user can interact with state-of-the-art rendering applications executed on the server. Our solution focuses on augmented video information, e.g., by depth, which is key to increase robustness with respect to data loss, image reconstruction,