鲁棒性压缩感知重构技术及其在智能视频监控中的应用研究

项目名称： 鲁棒性压缩感知重构技术及其在智能视频监控中的应用研究

项目编号： No.61501251

项目类型： 青年科学基金项目

立项/批准年度： 2016

项目学科： 无线电电子学、电信技术

项目作者： 杨真真

作者单位： 南京邮电大学

项目金额： 19万元

中文摘要： 压缩感知(Compressed Sensing, CS)是一种新兴的压缩采样技术，由于其具有对未知信号边采样边压缩的特性，所以该技术在许多领域特别是视频处理领域具有广阔的应用前景，它为智能视频监控系统所面临的“大数据”问题提供了一种新的解决方案。该技术中的核心问题是信号的稀疏性表示、压缩采样和重构，而通过求解非线性优化问题从少量观测中重构出稀疏信号的重构问题又起着举足轻重的作用。如何构造稳定、对观测数据要求少、计算复杂度低、收敛速度快的重构算法，一直是CS理论重构技术研究的主要内容和追求的目标。此外，在实际的应用中，噪声难以避免，所以研究压缩感知重构技术的鲁棒性，以改善重构信号质量，并将具有鲁棒性的CS技术应用到智能视频监控中，是十分重要和很有意义的课题。本课题的主要创新工作如下：①提出鲁棒性的视频联合重构技术；②提出基于鲁棒性压缩感知的目标检测算法；③提出基于鲁棒性压缩感知的图像融合算法

中文关键词： 压缩感知；鲁棒性；重构算法；目标检测；图像融合

英文摘要： Compressed sensing (CS) is an emerging compressive sampling technology and has attracted considerable attention in many fields especially the video processing field by suggesting that it samples a signal and compresses it meanwhile. It also provides a new solution to “big data” problem in intelligent video surveillance system. Compressed sensing contains three ideas: sparse representation, measurement and reconstruction, and one of the key ideas of it is to recover a sparse signal from very few measurements by nonlinear optimization. It is a goal of CS to tailor a stable, low computational complexity and fast convergence reconstruction algorithm. Besides，in actual applications, noises may inevitably exist, and thus to study the robustness of compressed sensing reconstruction technology is of great significance. It is important and necessary to tailor more robust reconstruction algorithms so as to improve the reconstruction performance, and apply the CS reconstruction technology to intelligent video surveillance. In this subject, the main contributions are as follows:①Tailor more robust video reconstruction technology; ②Tailor object detection algorithms based on robust CS; ③Tailor image fusion algorithms based on robust CS.

英文关键词： Compressed sensing;Robustness;Reconstruction algorithms;Object detection;Image fusion