We propose a full color video compression strategy, based on 3-D subband coding with camera pan compensation, to generate single embedded bit stream supporting multiple decoder display formats and wide, finely gradated range of rates. An experimental implementation our algorithm produces stream, from which suitable subsets are extracted be compatible many frame sizes rates satisfy transmission bandwidth constraints ranging several tens kilobits per second megabits second. Reconstructed...

The authors propose an iterative block reduction technique based on the theory of a projection onto convex sets. idea is to impose number constraints coded image in such way as restore it its original artifact-free form. One constraint can be derived by exploiting fact that transform-coded suffering from blocking effects contains high-frequency vertical and horizontal artifacts corresponding discontinuities across boundaries neighboring blocks. Another has with quantization intervals...

We introduce a point to real-time video transmission scheme over the Internet combining low-delay TCP-friendly transport protocol in conjunction with novel compression method that is error resilient and bandwidth-scalable. Compressed packetized into individually decodable packets of equal expected visual importance. Consequently, relatively constant quality can be achieved at receiver under lossy conditions. Furthermore, truncated instantaneously meet time varying bandwidth imposed by...

Managing multimedia data requires more than collecting the into storage archives and delivering it via networks to homes or offices. We survey technologies applications for video-content analysis retrieval. also give specific examples.

We present a video compression algorithm which performs well on generic sequences at very low bit rates. This was the basis for submission to November 1995 MPEG-4 subjective tests. The main novelty of is matching-pursuit based motion residual coder. method uses an inner-product search decompose signals overcomplete dictionary separable Gabor functions. coding strategy allows bits be concentrated in areas where they are needed most, providing detailed reconstructions without block artifacts....

The use of scalable video with layered multicast has been shown to be an effective method achieve rate control in heterogeneous networks. We propose the forward error correction (FEC) as error-control mechanism a framework. By organizing FEC into multiple layers, receivers can obtain different levels protection commensurate their respective channel conditions. Efficient network utilization is achieved streams are multicast, and only that need them. Furthermore, used without overall expansion...

The proliferation of video content on the Web makes similarity detection an indispensable tool in data management, searching, and navigation. We propose a number algorithms to efficiently measure similarity. define as set frames, which are represented high dimensional vectors feature space. Our goal is ideal (IVS), defined percentage clusters similar frames shared between two sequences. Since IVS too complex be deployed large database applications, we approximate it with Voronoi (VVS),...

One drawback of the discrete cosine transform (DCT) is visible block boundaries due to coarse quantization coefficients. Most restoration techniques for removing blocking effect are variations low-pass filtering, and as such, result in unnecessary blurring. The authors propose a new approach reducing which can be applied conventional coding without introducing additional information or significant method exploits correlation between intensity values boundary pixels two neighboring blocks. It...

In this letter, we present an approach to detecting trees in registered aerial image and range data obtained via lidar. The motivation for problem comes from automated 3-D city modeling, which such are used generate the models. Representing these models is problematic because usually too sparsely sampled tree regions create accurate model of trees. Furthermore, including points interferes with polygonization step building roof top Therefore, it advantageous detect remove that represent both...

We present an automatic system for planar 3D modeling of building interiors from point cloud data generated by range scanners. This is motivated the observation that most may be modeled as a collection planes representing ceilings, floors, walls and staircases. Our proposed system, which employs model-fitting RANSAC, capable detecting large-scale architectural structures, such ceilings well small-scale experimentally validate our on number challenging clouds real scenes.

Semantic understanding of environments is an important problem in robotics general and intelligent autonomous systems particular. In this paper, we propose a semantic segmentation algorithm which effectively fuses information from images 3D point clouds. The proposed method incorporates multiple scales intuitive effective manner. A late-fusion architecture to maximally leverage the training data each modality. Finally, pairwise Conditional Random Field (CRF) used as post-processing step...

Image based localization is an important problem with many applications. In our previous work, we presented a two step pipeline for performing image of mobile devices in outdoor environments. the first step, query matched against georeferenced 3D database to retrieve "closest" image. second pose recovered respect using cell phone sensors. As such, key ingredient database. this paper, extend approach indoors by utilizing locally referenced generated ambulatory depth acquisition backpack that...

We describe an approach to register and merge detailed facade models with a complementary airborne model. The modeling process provides half-meter resolution model bird's-eye view of the entire area, containing terrain profile building tops. ground-based results in facades. Using DSM obtained from laser scans, we localize acquisition vehicle facades by means Monte Carlo localization (MCL). two different resolutions obtain 3D

In this paper, we present a fast approach to automated generation of textured 3D city models with both high details at ground level, and complete coverage for bird's-eye view. A close-range facade model is acquired the level by driving vehicle equipped laser scanners digital camera under normal traffic conditions on public roads; far-range Digital Surface Map (DSM), containing complementary roof terrain shape, created from airborne scans, then triangulated, finally texture mapped aerial...

With the explosive growth of video applications over Internet, many approaches have been proposed to stream effectively packet switched, best-effort networks. A number these use techniques from source and channel coding, or implement transport protocols, modify system architectures in order deal with delay, loss, time-varying nature Internet. In this paper, we propose a framework for streaming multiple senders simultaneously single receiver. The main motivation doing so is exploit path...

In this paper we describe techniques for 3D textured model construction of urban areas using acquisition devices such as intensity cameras, well 2D laser scanner. Our experimental set up consists a truck equipped with one camera and two fast, inexpensive scanner, traveling on city streets under normal traffic conditions. The horizontal scans are used to determine the approximate component motion along movement vehicle. vertical scanner is build models facade buildings. To improve accuracy...

We propose an efficient bit allocation algorithm for a joint source/channel video codec over noisy channels. The approach is to distribute the available source and channel coding bits among subbands in such way that expected distortion minimized. constructed curves bound performance degradation should be estimated incorrectly. can used other similar minimization problems with two constraints, as power or complexity.

With the explosive growth of video applications over Internet, many approaches have been proposed to stream effectively packet switched, best-effort networks. We propose a receiver-driven protocol for simultaneous streaming from multiple senders single receiver in order achieve higher throughput, and increase tolerance loss delay due network congestion. Our employs novel rate allocation algorithm (RAA) partition (PPA). The RAA, run at receiver, determines sending each sender by taking into...

A framework for stability analysis of sigma-delta modulators is presented. It argued that limit cycles constant inputs are natural objects to investigate in this context. number analytical and approximate techniques aid the double loop interpolative discussed lead ways improving design explicitly take into account.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>

Packet loss and end-to-end delay limit sensitive applications over the best effort packet switched networks such as Internet. In our previous work, we have shown that substantial reduction in can be achieved by sending packets at appropriate rates to a receiver from multiple senders, using disjoint paths, protecting with forward error correction. this paper, propose path diversity correction (PDF) system for Internet which, paths sender are created collection of relay nodes. We scalable,...

3D modeling of building architecture from mobile scanning is a rapidly advancing field. These models are used in virtual reality, gaming, navigation, and simulation applications. State-of-the-art produces accurate point-clouds interiors containing hundreds millions points. This paper presents several scalable surface reconstruction techniques to generate watertight meshes that preserve sharp features the geometry common buildings. Our can automatically produce high-resolution fine detail...

Wireless multimedia delivery is becoming increasingly more important in today's networks. Unlike wired packet switched networks that suffer from congestion related loss and delay, the wireless have to deal with a time varying, error prone, physical channel many instances also severely bandwidth constrained. As such, solutions needed for video streaming applications are fundamentally different streaming. In this paper, we propose set of end application layer techniques adaptive over The...

This work describes an approach to texture mapping a 3D city model obtained from aerial and ground-based laser scans with oblique imagery. First, the images are automatically registered by matching 2D image lines projections of model. Then, for each triangle in model, optimal is selected taking into account occlusion, resolution, surface normal orientation, coherence neighboring triangles. Finally, utilized patches all combined one atlas compact representation efficient rendering. We...