TensorRT7 notes (two)

See an open source project on github: https://github.com/BaofengZan/DBNet-TensorRT

It is used to practice TensorRT.

First download the tensorRT7 library. This has been explained in Qianmian's notes (1). Here it is explained that I downloaded:

TensorRT-7.0.0.11.Windows10.x86_64.cuda-10.0.cudnn7.6

Correspondingly, I also installed cuda10.0 and cudnn7.6.5 locally.

Download the above project on Github to the local.

git clone https://github.com/BaofengZan/DBNet-TensorRT.git

Then create a new build folder under the DBNet-TensorRT folder, start powershell under the build folder, and execute:

cmake-gui ..

Configure the project through CMAKE-GUI.

The contents of the original project CmakeLists.txt are as follows (you need to set the opencv path and tensorrt path according to the local installation path when you use it yourself):

  
cmake_minimum_required(VERSION 2.6)

project(dbnet)

add_definitions(-std=c++11)

option(CUDA_USE_STATIC_CUDA_RUNTIME OFF)
set(CMAKE_CXX_STANDARD 11)
set(CMAKE_BUILD_TYPE Debug)

#设置cuda信息
find_package(CUDA REQUIRED)

message(STATUS "    libraries: ${CUDA_LIBRARIES}")
message(STATUS "    include path: ${CUDA_INCLUDE_DIRS}")
include_directories(${CUDA_INCLUDE_DIRS})
set(CUDA_NVCC_PLAGS ${CUDA_NVCC_PLAGS};-std=c++11; -g; -G;-gencode; arch=compute_75;code=sm_75)
enable_language(CUDA)  # 这一句添加后 ，就会在vs中不需要再手动设置cuda 

include_directories(${PROJECT_SOURCE_DIR}/include)

include_directories(D:\\TensorRT-7.0.0.11.Windows10.x86_64.cuda-10.2.cudnn7.6\\TensorRT-7.0.0.11\\include)

#-D_MWAITXINTRIN_H_INCLUDED  解决error: identifier "__builtin_ia32_mwaitx" is undefined
set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -std=c++11 -Wall -Ofast -D_MWAITXINTRIN_H_INCLUDED")

# 设置opencv的信息
set(OpenCV_DIR "D:\\opencv\\opencv346\\build")
find_package(OpenCV QUIET
    NO_MODULE
    NO_DEFAULT_PATH
    NO_CMAKE_PATH
    NO_CMAKE_ENVIRONMENT_PATH
    NO_SYSTEM_ENVIRONMENT_PATH
    NO_CMAKE_PACKAGE_REGISTRY
    NO_CMAKE_BUILDS_PATH
    NO_CMAKE_SYSTEM_PATH
    NO_CMAKE_SYSTEM_PACKAGE_REGISTRY
)

message(STATUS "OpenCV library status:")
message(STATUS "    version: ${OpenCV_VERSION}")
message(STATUS "    libraries: ${OpenCV_LIBS}")
message(STATUS "    include path: ${OpenCV_INCLUDE_DIRS}")

include_directories(${OpenCV_INCLUDE_DIRS})

link_directories(D:\\TensorRT-7.0.0.11.Windows10.x86_64.cuda-10.2.cudnn7.6\\TensorRT-7.0.0.11\\lib)

add_executable(dbnet ${PROJECT_SOURCE_DIR}/dbnet.cpp)

target_link_libraries(dbnet "nvinfer" "nvinfer_plugin" "nvparsers" "nvonnxparser")
#target_link_libraries(dbnet "cudart")
target_link_libraries(dbnet ${OpenCV_LIBS})
target_link_libraries(dbnet ${CUDA_LIBRARIES})

add_definitions(-pthread)

Note that you need to copy the tensorrt library files to the generated executable program directory. These files include (in addition to your own corresponding opencv dynamic library):

Change the following path to your own path

Execute first after compilation:

dbnet.exe -s

The DBNet.engine file will be generated in the current directory.

Then execute:

dbnet.exe -d ./test

The implementation effect is as follows:

More details will be explained in subsequent articles.

TensorRT7 notes (two)

Guess you like