基于ZYNQ UltraScale+MPSoC的实时视频图像锐化算法实现

Real-time Video Image Sharpening Algorithm Based on ZYNQ UltraScale+MPSoC is Implemented

本文通过为ZYNQ UltraScale+MPSoC移植Linux操作系统、交叉编译QT和OpenCV应用程序、外接一个USB摄像头、显示器,搭建了一套实时视频图像锐化处理系统。系统调用V4L2接口采集视频图像信息,然后将视频图像信息从内核空间映射到用户空间,便于应用程序快速读取、处理视频数据。另外通过QT应用程序调用OpenGL指令以驱动GPU采用3×3模板的Laplacian算子对采集到的视频图像进行迭代两次的锐化处理,最终将处理结果在显示器上显示。

This paper builds a real-time video image sharpening processing system by porting the Linux operating system,cross-compiling QT and OpenCV applications,and connecting a USB camera and display for ZYNQ UltraScale+MPSoC.The system calls the V4L2 interface to collect video image information,and then maps the video image information from the kernel space to the user space,which is convenient for applications to quickly read and process video data.In addition,the OpenGL instruction is called through the QT application to drive the GPU to use the Laplacian operator of the 3×3 template to sharpen the acquired video image twice,and finally display the processing result on the display.