基于Zynq的实时视频拼接技术研究与实现

发布时间：2018-01-25 18:15

本文关键词： Zynq 视频拼接图像拼接高层次综合软硬件协同设计　出处：《大连海事大学》2017年硕士论文　论文类型：学位论文

【摘要】：随着数字视频技术的快速发展,视频图像拼接技术成为计算机视觉领域重要的研究课题之一。目前,视频图像拼接技术在数字广播、视频监控、虚拟现实和医疗影像等领域得到广泛的应用。由于视频拼接技术要处理大量的图像数据,对处理速度要求高等问题,因此能够研究出一套实时性好、功耗低、体积小的视频拼接系统具有重要的实际意义。本系统采用的是Xilinx公司Zynq-7000系列全面可编程片上系统平台,片内集成了双核ARM Cortex-A9处理器和Xilinx 7系列FPGA,且内部有高速互联结构,采用软硬件协同设计方法,将图像采集、视频拼接、高清显示集合到一个嵌入式系统中。与单一处理器如DSP、ARM、FPGA相比,这种方法既充分利用了 FPGA强大的并行运算能力与丰富的逻辑资源,又结合了 ARM在搭建操作系统与实现复杂算法方面的优点。本文主要工作有:(1)在搭载Zynq芯片的ZedBoard开发板上,构建嵌入式Linux开发环境,包括启动文件的制作,交叉编译环境的搭建,OpenCV和Qt的安装与移植。(2)利用Vivado高层次综合工具(High-Level Synthesis,HLS)设计自定义硬件加速IP核,包括Harris角点检测算法和图像傅里叶变换的倒位序算法。(3)在Zynq芯片的可编程逻辑(Programmable Logic,PL)部分完成硬件工程的搭建,主要包括硬件加速IP核的挂载与配置、视频图像的直接存储器存取(Video Direct Memory Access,VDMA)配置以及 HDMI 高清显示控制。(4)在Zynq芯片的处理器系统(Processing System,PS)部分即ARM内实现系统软件设计,主要包括VDMA IP核和自定义硬件加速IP核的驱动程序设计,以及使用OpenCV函数库和自定义硬件加速IP核实现视频图像采集以及视频拼接功能。本设计采用软硬件协同设计方法完成了基于Zynq实时视频拼接系统,具有开发周期短、实时性好、体积小、界面友好等特点。
[Abstract]:With the rapid development of digital video technology, video image mosaic technology has become one of the important research topics in the field of computer vision. At present, video image mosaic technology in digital broadcasting, video surveillance. Virtual reality and medical images have been widely used. Because video mosaic technology has to deal with a large number of image data, high speed of processing problems, so it can develop a set of real-time, low power consumption. The small size of the video splicing system is of great practical significance. This system uses the Xilinx Zynq-7000 series of comprehensive programmable chip system platform. The dual-core ARM Cortex-A9 processor and the Xilinx 7 series FPGA are integrated in the chip, and there is a high speed interconnection structure inside. The image is collected by using the method of hardware and software co-design. Video splicing, high-definition display set into an embedded system, compared with a single processor such as DSP ARMU FPGA. This method not only makes full use of FPGA's powerful parallel computing ability and rich logical resources. Combined with the advantages of ARM in building the operating system and implementing complex algorithms. The main work of this paper is: 1) on the ZedBoard development board with Zynq chip. Build the embedded Linux development environment, including the production of startup files, cross-build the environment. Installation and migration of OpenCV and QT. 2) make use of the Vivado high-level Synthesis. HLS) designs custom hardware-accelerated IP cores. It includes Harris corner detection algorithm and image Fourier transform inversion algorithm. 3) Programmable Logic in Zynq chip. PL) part completes the hardware engineering construction, mainly includes the hardware acceleration IP core mount and the configuration. Video Direct Memory Access is accessed by direct memory of video image. VDMA) configuration and HDMI HD display control. 4) processing System in the processor system of the Zynq chip. PS) part is the software design of the system in ARM, mainly including the VDMA IP core and the driver design of the custom hardware accelerated IP core. The function of video image acquisition and video splicing is realized by using OpenCV function library and self-defined hardware accelerated IP core. This design adopts the method of hardware and software co-design to complete real-time video mosaic based on Zynq. The system. Has the development cycle short, the real-time good, the volume is small, the interface friendly and so on characteristic.
【学位授予单位】：大连海事大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP391.41

【相似文献】