ITR/SI(SPS)+AP+SY: High Resolution Video and Still Reconstruction with Applications to Visual Communications and Image Enhancement

高分辨率高精度视频重建及其在视觉通信和图像增强中的应用压缩数据的分辨率和比特深度增强问题对于新兴的、易出错的、有限带宽的无线和异构计算环境中的应用将变得越来越重要。该研究涉及通过多帧和多通道处理来提高视频信号的空间、时间和比特深度分辨率。作为这项工作的一部分开发的系统适用于视频通信和多媒体处理的各个领域。分配用于表示图像中点处的强度或颜色信息的比特数通常被称为比特深度。当位深度不够时，图像会出现脊状结构，称为假轮廓。当需要低对比度细节时，位深度限制变得很重要，如在医学成像、航空/卫星摄影和高质量扫描应用中。注意到比特深度和空间分辨率增强之间的相似性，本研究探讨了一种增加比特深度的方法。 具体地说，当一个序列的视频帧是可用的，它是可能的，以实现更高的比特深度和更高的空间分辨率，通过投影到凸集（POCS）为基础的数据融合method.The多帧视频处理算法和它们的多媒体和通信应用形成的信号处理教育议程的一个组成部分，在格鲁吉亚技术。 在多帧处理（学生）的兴趣水平的激励和鼓励下，该项目构建了一个以网络为中心的图像分辨率和精度增强演示，以造福于本科生和研究社区。

ABSTRACTProposal 0113681Altunbasak, YucelGA Tech Res Corp - GITITR/SI(SPS)+AP+SY: High-Resolution High-Precision Video Reconstructionwith Applications to Visual Communications and Image EnhancementThe problem of resolution and bit-depth enhancement from compressed data will become increasingly important for applications in the emerging, error-prone, restricted bandwidth, wireless and heterogeneous computing environments. This research involves enhancing the spatial, temporal and bit-depth resolution of video signals by means of multi-frame and multi-channel processing. The system that is developed as part of this work is applicable to various areas of video communications and multi-media processing.The number of bits assigned to represent the intensity or color information at points in an image is usually referred as the bit depth. When the bit depth is not sufficient, images suffer from ridge-like structures known as false contours. Bit-depth limitations become important when low-contrast details are required, as in medical imaging, aerial/satellite photography, and high-quality scanning applications. Noting the similarity between the bit-depth and spatial resolution enhancement, this research examines a method for increasing the bit depth. Specifically, when a sequence of video frames is available, it is possible to achieve a higher bit depth and higher spatial resolution through a projections-onto-convex-sets (POCS) based data fusion method.The multiple-frame video processing algorithms and their multimedia and communication applications form an integral part of the signal processing education agenda at Georgia Tech. Motivated and encouraged by the level of interest in multi-frame processing (by the students), the project constructs a web-centric demo of image resolution and precision enhancement for the benefit of the undergraduates as well as the research community.