基于压缩感知的ARM嵌入式视频监控系统

发布时间：2018-01-06 10:01

本文关键词：基于压缩感知的ARM嵌入式视频监控系统　出处：《太原科技大学》2012年硕士论文　论文类型：学位论文

【摘要】：目前，视频监控以其直观、方便、内容信息丰富而被广泛应用于交通、城市安防、工业、国防等多种领域。传统的视频监控系统基于奈奎斯特采样定理进行视频信号的采集，然后采用MPEG、H.26X系列算法对视频图像进行压缩，导致信息采集数据量大，视频压缩算法复杂，给视频信号存储、压缩和传输带来了困难。另一方面，传统的视频监控系统大多采用在PC机上安装视频采集卡和通信接口来进行视频信号的压缩和传输，存在体积庞大，结构复杂，价格昂贵，施工布线复杂等缺点。对此，本文在深入研究压缩感知视频编解码算法，ARM嵌入式体系结构和Linux操作系统移植的基础上，将压缩感知视频编码算法与ARM嵌入式开发技术相结合，，实现了基于压缩感知的ARM嵌入式视频监控系统，一定程度上克服了传统视频监控带来的缺陷。本论文主要工作和成果如下： 1）详细研究并讨论了具有低复杂度编码特性的分布式压缩感知理论基础，在此基础上提出了基于残差分布式视频压缩感知算法，并用实验仿真验证了算法的有效性。 2）深入研究了ARM嵌入式平台的软硬件技术，将压缩感知算法与ARM嵌入式相结合，设计了基于压缩感知算法的ARM嵌入式视频监控系统硬件架构和软件平台。 3）详细阐述了压缩感知视频监控系统中的关键技术，包括压缩感知USB摄像头图像采集系统和驱动模块软件设计，V4L2视频图像采集模块设计，压缩感知视频编码算法优化和移植，本地视频监控存储实现。
[Abstract]:At present, the video monitor with its intuitive, convenient, content rich information and is widely used in city traffic, security, industrial, defense and other fields. The video surveillance system of traditional Nyquist sampling theorem for video signal acquisition based on, and then using MPEG, H.26X series algorithm to compress the video image, resulting in information collection data large, complex video compression algorithm for video signal storage, compression and transmission difficult. On the other hand, the traditional video monitoring system mostly adopts the installation of video capture card and communication interface in the PC machine to carry out video signal compression and transmission, huge volume, complicated structure, expensive, complex wiring construction defects in this regard, based on the in-depth study of compressed sensing video codec algorithm, ARM based transplantation of embedded system and Linux operating system, the compressed sensing video encoding Combined with ARM embedded development technology, the ARM embedded video surveillance system based on compressed sensing is implemented, which overcomes the shortcomings of traditional video surveillance to a certain extent.
The main work and results of this paper are as follows:
1) the theory of distributed compressed sensing with low complexity coding is studied and discussed in detail. Based on it, a distributed video compressed sensing algorithm based on residual is proposed, and the effectiveness of the algorithm is verified by experimental simulation.
2) deeply studied the software and hardware technology of ARM embedded platform, combined the compressed sensing algorithm with ARM embedded technology, designed the hardware architecture and software platform of ARM embedded video monitoring system based on compressed sensing algorithm.
3) details the key technology of compressed sensing in video monitoring system, including the perception of USB camera image acquisition system and module software design, design of video image acquisition module V4L2, compressed sensing video encoding optimization algorithm and transplantation, local video surveillance storage.

【学位授予单位】：太原科技大学
【学位级别】：硕士
【学位授予年份】：2012
【分类号】：TP277;TP368.1

【参考文献】