如何在Visual Studio 2013中用CMake构建多源文件PCL项目

Hey there! Let's tackle your questions about linking a static library to your main program, plus share some solid learning resources for CMake, VS2013, and PCL—super excited to help with this cool robotic 3D registration project!

1. Yes, You Need Corresponding Header Files #

Each of your library source files (like xyz2pcd.cpp, statistical_removal.cpp) requires a matching header file (e.g., xyz2pcd.h, statistical_removal.h) to expose the functions/classes you want to call from project_test.cpp.

For example, if xyz2pcd.cpp has a function like:

void convertXYZToPCD(const std::string& xyz_path, const std::string& pcd_path) {
    // Your conversion logic here
}

Your xyz2pcd.h should declare this function (with include guards to avoid duplicate definitions):

#ifndef XYZ2PCD_H
#define XYZ2PCD_H

#include <string>

void convertXYZToPCD(const std::string& xyz_path, const std::string& pcd_path);

#endif // XYZ2PCD_H

Repeat this pattern for all your library files—only expose the functions/classes your main program needs to use.

2. Adjust Your CMakeLists.txt to Link the Library #

Here's a complete, tailored CMakeLists.txt example that compiles your library sources into a static library, then links it to your main executable along with PCL dependencies:

cmake_minimum_required(VERSION 3.8.0-rc1)
project(Robotic3DRegistration)

# Find required PCL components
find_package(PCL 1.8.0 REQUIRED COMPONENTS common io filters registration features)
include_directories(${PCL_INCLUDE_DIRS})
link_directories(${PCL_LIBRARY_DIRS})
add_definitions(${PCL_DEFINITIONS})

# Compile your static library
add_library(project_lib STATIC
    xyz2pcd.cpp
    statistical_removal.cpp
    moment_of_inertia.cpp
    pairwise_incremental_registration.cpp
    robot_control.cpp
    sensor_control.cpp
)

# Link PCL libraries to your static library (since it uses PCL functions)
target_link_libraries(project_lib ${PCL_LIBRARIES})

# Build your main executable
add_executable(project_test project_test.cpp)

# Link your static library and PCL to the main executable
target_link_libraries(project_test project_lib ${PCL_LIBRARIES})

Key details:

• add_library(project_lib STATIC ...) creates the static library from your source files.
• target_link_libraries(project_lib ${PCL_LIBRARIES}) ensures your library has access to PCL's functionality.
• The final target_link_libraries connects your main program to both your custom library and PCL.

3. Calling Library Functions in project_test.cpp #

Once headers and CMake are set up, using the library functions is straightforward:

#include "xyz2pcd.h"
#include "statistical_removal.h"
// Include other required headers

int main() {
    // Call your library functions directly
    convertXYZToPCD("scan_data.xyz", "scan_data.pcd");
    
    pcl::PointCloud<pcl::PointXYZ>::Ptr cloud(new pcl::PointCloud<pcl::PointXYZ>);
    // Load your point cloud...
    removeStatisticalOutliers(cloud); // Example function from statistical_removal.cpp
    
    // Rest of your multi-view registration and robot control logic
    return 0;
}

4. Learning Resources for CMake, VS2013, and PCL #

CMake #

• Mastering CMake: Kitware's official book covers core concepts like library building, target linking, and VS-specific generator setup—perfect for getting a deep understanding.
• CMake Official Tutorials: Focus on sections about add_library, target_link_libraries, and C++ project workflows for Visual Studio.
• Small-scale practice: Start with a simple 2-file library + executable project to build confidence before scaling to your full project.

Visual Studio 2013 + PCL #

• PCL Official Docs: Check the "Building PCL Applications" section for VS-specific tips on integrating CMake-generated projects and PCL dependencies.
• VS2013 C++ Project Guide: Microsoft's official docs for managing multi-project solutions, linking libraries, and debugging C++ code (stick to 2013-specific content since you're using an older version).
• Open-source PCL Projects: Browse GitHub for PCL-based robotic projects—study their CMakeLists.txt files and project structures to pick up best practices.

PCL Specific #

• Point Cloud Library: From Algorithms to Applications: A comprehensive book covering PCL modules like registration, filtering, and ICP, with code examples directly applicable to your multi-view reconstruction work.
• PCL Official Tutorials: The "Registration" and "Multi-View Reconstruction" sections are directly relevant—they walk through ICP implementation and multi-view point cloud processing workflows.

内容的提问来源于stack exchange，提问作者user9862453