一、前言
想在Xilinx ZCU102平台上運作Linux系統，并實作簡單OpenCV App的運作，采用Petalinux19.1進行工程搭建。
二、OpenCV App搭建
1.準備Petalinux工程
1）建立Petalinux工程：
>petalinux-create --type project --template zynqMP --name opencvtest
2）指定HDF檔案：
>petalinux-config --get-hw-description=hdf_file/
3）使能OpenCV、X11的Package：
>petalinux-config -c rootfs
使能
使能
4）Build工程，封裝得到BOOT.BIN和image.ub
>petalinux-build
>petalinux-package --boot --fsbl images/linux/zynqmp_fsbl.elf --fpga images/linux/system.bit --u-boot images/linux/u-boot.elf --force
把BOOT.BIN和image.ub考入SD卡，ZCU102平台可運作Linux；
2.建立OpenCV App
>petalinux--create -t apps --template c++ --name opencvapp --enable
>cd xxx/project-spec/meta-user/recipes-apps/opencvapp/files
1)opencvapp.cpp
#include <stdio.h>
#include <opencv2/highgui/highgui.hpp>
using namespace cv;
using namespace std;
int main(int argc, char** argv){
Mat img;
img = imread(argv[1],1);
if((argc != 2) || (!img.data)){
printf("Usage: XX yy.jpg\n");
return -1;
}
imshow("imshooow",img);
waitKey(0);
return 0;
}
2）CMakeLists.txt
# cmake needs this line
cmake_minimum_required(VERSION 2.8)
add_definitions(-std=c++11 -Werror=return-type)
# Define project name
project(opencvapp)
# Find OpenCV, you may need to set OpenCV_DIR variable
# to the absolute path to the directory containing OpenCVConfig.cmake file
# via the command line or GUI
find_package(OpenCV REQUIRED)
if(CMAKE_VERSION VERSION_LESS "2.8.11")
# Add OpenCV headers location to your include paths
include_directories(${OpenCV_INCLUDE_DIRS})
endif()
set(SRC_LIST opencvapp.cpp )
# Declare the executable target built from your sources
add_executable(opencvapp ${SRC_LIST})
# Link your application with OpenCV libraries
target_link_libraries(opencvapp ${OpenCV_LIBS})
3）Makefile
APP = opencvapp
# Add any other object files to this list below
APP_OBJS = opencvapp.o
all: build
build: $(APP)
CXXFLAGS += $(shell pkg-config --cflags opencv)
LDFLAGS += $(shell pkg-config --libs opencv)
$(APP): $(APP_OBJS)
$(CXX) $(CXXFLAGS) $(LDFLAGS) -o [email protected] $(APP_OBJS) $(LDLIBS)
clean:
rm -f $(APP) *.elf *.gdb *.o
4) opencvapp.bb
#
# This file is the opencvapp recipe.
#
SUMMARY = "Simple opencvapp application"
SECTION = "PETALINUX/apps"
LICENSE = "MIT"
LIC_FILES_CHKSUM = "file://${COMMON_LICENSE_DIR}/MIT;md5=0835ade698e0bcf8506ecda2f7b4f302"
SRC_URI = "file://opencvapp.cpp \
file://CMakeLists.txt \
"
S = "${WORKDIR}"
DEPENDS += "opencv"
inherit pkgconfig cmake
do_install() {
install -d ${D}${bindir}
install -m 0755 opencvapp ${D}${bindir}
}
3. 編譯App
直接編譯app：
>petalinux-build -c opencvapp
編譯完成後，找到xxxxx/tmp/opencvtest-2019.12.xx/deploy/rpm/aarch64/opencvapp-1.0-r0.aarch64.rpm檔案，
将其在windows下解壓，再解壓，可得到：
/usr/bin/opencvapp
4. 運作
将opencvapp和test.jpg經SSH傳入已運作的ZCU102平台中，
>chmod 777 opencvapp test.jpg
配置SSH顯示到本地：export DISPLAY=192.168.0.xx:0.0，然後運作
>./opencvapp test.jpg
可（經SSH）顯示出圖檔。