From aaed54421889392d832581a6b0e2d3014aa19ab8 Mon Sep 17 00:00:00 2001 From: zhaoxi <535394140@qq.com> Date: Tue, 4 Jun 2024 01:02:53 +0800 Subject: [PATCH] =?UTF-8?q?=E6=96=B0=E7=9A=84=20ort=20=E5=A4=9A=E6=80=81?= =?UTF-8?q?=E9=83=A8=E7=BD=B2=E5=BA=93=EF=BC=8C=E4=BB=A5=E6=94=AF=E6=8C=81?= =?UTF-8?q?=20CUDA=E3=80=81TensorRT=20=E5=92=8C=20OpenVINO=20=E7=9A=84=20P?= =?UTF-8?q?rovider=20=E9=80=89=E6=8B=A9?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit --- CMakeLists.txt | 6 +- README.md | 12 +- cmake/RMVLModule.cmake | 3 - doc/tutorials/introduction/linux/install.md | 13 +- doc/tutorials/modules/algorithm/lsqnonlin.md | 4 +- doc/tutorials/modules/tools/opcua.md | 14 +- extra/detector/CMakeLists.txt | 2 +- .../include/rmvl/detector/armor_detector.h | 15 +- .../detector/include/rmvl/detector/detector.h | 2 +- .../include/rmvl/detector/gyro_detector.h | 17 +- extra/detector/src/armor_detector/detect.cpp | 2 +- extra/detector/src/armor_detector/find.cpp | 7 +- extra/detector/src/gyro_detector/find.cpp | 9 +- extra/detector/src/gyro_detector/match.cpp | 8 +- include/rmvl/rmvl.hpp | 4 - .../camera/include/rmvl/camera/camutils.hpp | 2 +- modules/core/CMakeLists.txt | 2 +- .../include/rmvl/core/pretreat.hpp} | 22 +- .../{imgproc => core}/perf/perf_pretreat.cpp | 14 +- modules/{imgproc => core}/src/pretreat.cpp | 2 +- modules/core/test/test_optimal.cpp | 2 +- .../{imgproc => core}/test/test_pretreat.cpp | 14 +- modules/imgproc/CMakeLists.txt | 23 -- modules/imgproc/include/rmvl/imgproc.hpp | 19 -- .../imgproc/include/rmvl/imgproc/paint.hpp | 175 -------------- modules/imgproc/test/test_paint.cpp | 65 ------ modules/ml/include/rmvl/ml.hpp | 10 +- modules/ml/include/rmvl/ml/ort.h | 141 +++++++++-- modules/ml/src/ort/classification.cpp | 110 +++++++++ modules/ml/src/ort/ort.cpp | 220 +++++++++--------- modules/ml/src/ort/ort_impl.h | 66 ------ modules/ml/src/ort/ort_print.cpp | 43 ---- modules/opcua/src/client.cpp | 18 +- modules/opcua/src/helper.cpp | 1 + modules/opcua/src/server.cpp | 27 ++- modules/opcua/test/test_opcua_client.cpp | 12 +- samples/opcua/opcua_client.cpp | 2 +- 37 files changed, 474 insertions(+), 634 deletions(-) rename modules/{imgproc/include/rmvl/imgproc/pretreat.h => core/include/rmvl/core/pretreat.hpp} (73%) rename modules/{imgproc => core}/perf/perf_pretreat.cpp (93%) rename modules/{imgproc => core}/src/pretreat.cpp (97%) rename modules/{imgproc => core}/test/test_pretreat.cpp (91%) delete mode 100644 modules/imgproc/CMakeLists.txt delete mode 100644 modules/imgproc/include/rmvl/imgproc.hpp delete mode 100644 modules/imgproc/include/rmvl/imgproc/paint.hpp delete mode 100644 modules/imgproc/test/test_paint.cpp create mode 100644 modules/ml/src/ort/classification.cpp delete mode 100644 modules/ml/src/ort/ort_impl.h delete mode 100755 modules/ml/src/ort/ort_print.cpp diff --git a/CMakeLists.txt b/CMakeLists.txt index 5000586b..b1f0b677 100755 --- a/CMakeLists.txt +++ b/CMakeLists.txt @@ -195,7 +195,11 @@ foreach(m ${RMVL_EXTRA_MODULE}) if("${m}" STREQUAL "types") continue() endif() - string(REGEX REPLACE "rmvl_" " " ${m}_modules ${RMVL_${mUP}_MODULES_BUILD}) + if("${RMVL_${mUP}_MODULES_BUILD}" STREQUAL "") + set(${m}_modules "") + else() + string(REGEX REPLACE "rmvl_" " " ${m}_modules ${RMVL_${mUP}_MODULES_BUILD}) + endif() endforeach() string(REGEX REPLACE "rmvl_" " " main_modules ${RMVL_MAIN_MODULES_BUILD}) diff --git a/README.md b/README.md index d174c6ea..33de3b11 100755 --- a/README.md +++ b/README.md @@ -2,17 +2,17 @@ **机器人控制与视觉库** -| 构建配置 | 编译器/环境 | Github Actions 工作流状态 | -| :------: | :-------------------------------: | :--------------------------------------: | -| CMake | GCC 7.5.0 x86_64-linux-gnu
GCC 9.4.0 x86_64-linux-gnu
GCC 12.3.0 x86_64-linux-gnu | [![1.x in Linux](https://github.com/cv-rmvl/rmvl/actions/workflows/linux-1.x.yml/badge.svg)](https://github.com/cv-rmvl/rmvl/actions/workflows/linux-1.x.yml) | +| 构建配置 | 编译器/环境 | Github Actions 工作流状态 | +| :------: | :----------------------------------------------------------: | :----------------------------------------------------------: | +| CMake | GCC 7.5.0 x86_64-linux-gnu
GCC 9.4.0 x86_64-linux-gnu
GCC 12.3.0 x86_64-linux-gnu
GCC 14.0.1 x86_64-linux-gnu | [![1.x in Linux](https://github.com/cv-rmvl/rmvl/actions/workflows/linux-1.x.yml/badge.svg)](https://github.com/cv-rmvl/rmvl/actions/workflows/linux-1.x.yml) | -RMVL 最初是面向 RoboMaster 赛事的视觉库，因此称为 RoboMaster Vision Library，现计划并逐步完善与机器人相关的视觉、控制、通信的功能，旨在打造适用范围广、使用简洁、架构统一、功能强大的视觉控制一体库。 +RMVL 最初是面向 RoboMaster 赛事的视觉库，现在此之上逐步完善有关基础算法、机器视觉、通信的功能，旨在打造适用范围广、使用简洁、架构统一、功能强大的视觉控制一体库。 ### 项目资源 -* 说明文档: +* 用户手册: -* 工单: +* 问题: * VSCode 扩展插件 diff --git a/cmake/RMVLModule.cmake b/cmake/RMVLModule.cmake index ef89b1b5..295ffdf0 100644 --- a/cmake/RMVLModule.cmake +++ b/cmake/RMVLModule.cmake @@ -100,9 +100,6 @@ macro(rmvl_add_module _name) if(NOT DEFINED BUILD_${the_module}_INIT) set(BUILD_${the_module}_INIT ON) endif() - if(DEFINED BUILD_${the_module}) - unset(BUILD_${the_module} CACHE) - endif() option(BUILD_${the_module} "Include ${the_module} module into the RMVL build" ${BUILD_${the_module}_INIT}) # create option to enable/disable this module if(BUILD_${the_module}) diff --git a/doc/tutorials/introduction/linux/install.md b/doc/tutorials/introduction/linux/install.md index a19ddc98..c48645d9 100644 --- a/doc/tutorials/introduction/linux/install.md +++ b/doc/tutorials/introduction/linux/install.md @@ -130,9 +130,7 @@ sudo apt install libeigen3-dev #### 1.4 onnxruntime{#install_onnxruntime} -onnxruntime 库是目前数字识别所依赖的第三方库，如果有需要开启此功能，则需要安装 onnxruntime - -- 获取压缩包 +- 获取压缩包，可选择对应的发行版，也可以是 gpu、cuda12 或 cpu 版本，下面以 1.12.0 的 cpu 版本为例 ```shell curl -SL https://github.com/microsoft/onnxruntime/releases/download/v1.12.0/onnxruntime-linux-x64-1.12.0.tgz -o onnxruntime-linux-x64-1.12.0.tgz ``` @@ -143,12 +141,9 @@ onnxruntime 库是目前数字识别所依赖的第三方库，如果有需要 - 安装（复制头文件与库文件） ```shell sudo mkdir /usr/local/include/onnxruntime - sudo cp onnxruntime-linux-x64-1.12.0/include/* /usr/local/include/onnxruntime - sudo cp -r onnxruntime-linux-x64-1.12.0/lib /usr/local - ``` -- 移除中间文件 - ```shell - rm -r onnxruntime-linux-x64-1.12.0 && rm onnxruntime-linux-x64-1.12.0.tgz + cd onnxruntime-linux-x64-1.12.0 + sudo cp -r include/* /usr/local/include/onnxruntime + sudo cp -r lib /usr/local ``` 至此 onnxruntime 安装完成。 diff --git a/doc/tutorials/modules/algorithm/lsqnonlin.md b/doc/tutorials/modules/algorithm/lsqnonlin.md index 258625a3..31c7b9df 100644 --- a/doc/tutorials/modules/algorithm/lsqnonlin.md +++ b/doc/tutorials/modules/algorithm/lsqnonlin.md @@ -72,7 +72,7 @@ 1. 选择一初始点 \f$\pmb x_0=(x_{1,0},x_{2,0},\cdots,x_{n,0})^T\f$； 2. 算出 \f[\Delta\pmb x_0=-\pmb H_0^{-1}\pmb J_0^T\pmb\varphi(\pmb x_k)\tag{1-6}\f] 3. 令 \f$\pmb x_1\f$ 为函数 \f$f(\pmb x)\f$ 的极小点的第 1 次近似，则有 \f[\pmb x_1=\pmb x_0+\Delta\pmb x_0\tag{1-7}\f] -4. 以 \f$\pmb x_1\f$ 代替前面的 \f$\pmb x_0\f$，\f$\Delta\pmb x_1\f$ 代替 \f$\Delta\pmb x_0\f$，重复上述计算过程，直到 \f[\|\pmb\varphi(\pmb x_k)\|<\epsilon'\tag{1-8a}\f] 或 \f[\|\nabla f(\pmb x_k)\|<\epsilon"\tag{1-8b}\f] 为止。\f$\epsilon'\f$ 和 \f$\epsilon"\f$ 是预先给定的精度。 +4. 以 \f$\pmb x_1\f$ 代替前面的 \f$\pmb x_0\f$，\f$\Delta\pmb x_1\f$ 代替 \f$\Delta\pmb x_0\f$，重复上述计算过程，直到 \f[\|\pmb\varphi(\pmb x_k)\|<\epsilon'\tag{1-8a}\f] 或 \f[\|\nabla f(\pmb x_k)\|<\epsilon''\tag{1-8b}\f] 为止。\f$\epsilon'\f$ 和 \f$\epsilon''\f$ 是预先给定的精度。 #### 1.3 改进 @@ -85,7 +85,7 @@ #### 1.4 如何使用 -可参考 `rm::lsqnonlin` 函数，例如，我们需要拟合一个正弦函数，下面的 `obtain` 函数就是观测每一帧的数据。 +RMVL 提供了改进的 Gauss-Newton 迭代算法，可参考 `rm::lsqnonlin` 函数。例如，我们需要拟合一个正弦函数\f[y=A\sin(\omega t+\varphi_0)+b\f]其中，\f$A,\omega,\varphi_0,b\f$ 是待拟合的参数，不妨统一写为 \f$\pmb x=(A,\omega,\varphi_0,b)\f$，也就是说我们需要拟合的函数是\f[\green y=x_1\sin(x_2\green t+x_3)+x_4\f]其中 \f$t\f$ 和 \f$y\f$ 是可以观测到的数据，我们需要通过观测的数据来拟合 \f$\pmb x\f$ 的值。比方说，下面的 `obtain` 函数就可以观测每一帧的数据。 ```cpp double obtain(); diff --git a/doc/tutorials/modules/tools/opcua.md b/doc/tutorials/modules/tools/opcua.md index b0c63aa0..ea8eb654 100644 --- a/doc/tutorials/modules/tools/opcua.md +++ b/doc/tutorials/modules/tools/opcua.md @@ -97,8 +97,8 @@ int main() int main() { - // 创建 OPC UA 客户端，连接到 localhost:4840 - rm::Client clt("opc.tcp://localhost:4840"); + // 创建 OPC UA 客户端，连接到 127.0.0.1:4840 + rm::Client clt("opc.tcp://127.0.0.1:4840"); /* other code */ } @@ -141,7 +141,7 @@ int main() int main() { - rm::Client clt("opc.tcp://localhost:4840"); + rm::Client clt("opc.tcp://127.0.0.1:4840"); // 使用管道运算符 "|" 进行路径搜索，寻找待读取的变量 auto node = rm::nodeObjectsFolder | clt.find("number"); @@ -210,7 +210,7 @@ int main() int main() { - rm::Client clt("opc.tcp://localhost:4840"); + rm::Client clt("opc.tcp://127.0.0.1:4840"); // 设置输入参数，1 和 2 是 Int32 类型的，因此可以直接隐式构造 std::vector input = {1, 2}; @@ -287,7 +287,7 @@ int main() int main() { - rm::Client clt("opc.tcp://localhost:4840"); + rm::Client clt("opc.tcp://127.0.0.1:4840"); // 路径搜索寻找 C2 auto node_c2 = rm::nodeObjectsFolder | clt.find("A") | clt.find("B1") | clt.find("C2"); @@ -369,7 +369,7 @@ int main() int main() { - rm::Client clt("opc.tcp://localhost:4840"); + rm::Client clt("opc.tcp://127.0.0.1:4840"); auto node = rm::nodeObjectsFolder | clt.find("number"); for (int i = 0; i < 100; ++i) { @@ -396,7 +396,7 @@ void onChange(UA_Client *, UA_UInt32, void *, UA_UInt32, void *, UA_DataValue *v int main() { - rm::Client clt("opc.tcp://localhost:4840"); + rm::Client clt("opc.tcp://127.0.0.1:4840"); auto node = rm::nodeObjectsFolder | clt.find("number"); // 监视变量，这里的 onChange 同样可以写成无捕获列表的 lambda 表达式，因为存在隐式转换 client.monitor(node_id, onChange, 5); diff --git a/extra/detector/CMakeLists.txt b/extra/detector/CMakeLists.txt index a306fd74..20aff70f 100755 --- a/extra/detector/CMakeLists.txt +++ b/extra/detector/CMakeLists.txt @@ -3,7 +3,7 @@ # ---------------------------------------------------------------------------- rmvl_add_module( detector INTERFACE - DEPENDS imgproc group + DEPENDS group ) # armor_detector diff --git a/extra/detector/include/rmvl/detector/armor_detector.h b/extra/detector/include/rmvl/detector/armor_detector.h index 4a2206c2..c115b6c1 100755 --- a/extra/detector/include/rmvl/detector/armor_detector.h +++ b/extra/detector/include/rmvl/detector/armor_detector.h @@ -28,7 +28,7 @@ namespace rm //! 装甲板识别模块 class ArmorDetector final : public detector { - std::unique_ptr _ort; + std::unique_ptr _ort; std::unordered_map _robot_t; public: @@ -37,16 +37,9 @@ class ArmorDetector final : public detector explicit ArmorDetector(const std::string &model) { - _ort = std::make_unique(model); - _robot_t[0] = RobotType::UNKNOWN; - _robot_t[1] = RobotType::HERO; - _robot_t[2] = RobotType::ENGINEER; - _robot_t[3] = RobotType::INFANTRY_3; - _robot_t[4] = RobotType::INFANTRY_4; - _robot_t[5] = RobotType::INFANTRY_5; - _robot_t[6] = RobotType::OUTPOST; - _robot_t[7] = RobotType::BASE; - _robot_t[8] = RobotType::SENTRY; + _ort = std::make_unique(model); + for (int i = 0; i < 9; ++i) + _robot_t[i] = static_cast(i); } /** diff --git a/extra/detector/include/rmvl/detector/detector.h b/extra/detector/include/rmvl/detector/detector.h index 918ced81..c9532e8e 100755 --- a/extra/detector/include/rmvl/detector/detector.h +++ b/extra/detector/include/rmvl/detector/detector.h @@ -13,7 +13,7 @@ #include "rmvl/group/group.h" -#include "rmvl/imgproc/pretreat.h" +#include "rmvl/core/pretreat.hpp" namespace rm { diff --git a/extra/detector/include/rmvl/detector/gyro_detector.h b/extra/detector/include/rmvl/detector/gyro_detector.h index 6c58f82f..4d91d75e 100644 --- a/extra/detector/include/rmvl/detector/gyro_detector.h +++ b/extra/detector/include/rmvl/detector/gyro_detector.h @@ -29,25 +29,18 @@ class GyroDetector final : public detector { int _armor_num; //!< 默认装甲板数目 - std::unique_ptr _ort; + std::unique_ptr _ort; std::unordered_map _robot_t; public: GyroDetector(int armor_num) : _armor_num(armor_num) {} ~GyroDetector() = default; - explicit GyroDetector(const std::string &model, int armor_num) : _armor_num(armor_num) + GyroDetector(const std::string &model, int armor_num) : _armor_num(armor_num) { - _ort = std::make_unique(model); - _robot_t[0] = RobotType::UNKNOWN; - _robot_t[1] = RobotType::HERO; - _robot_t[2] = RobotType::ENGINEER; - _robot_t[3] = RobotType::INFANTRY_3; - _robot_t[4] = RobotType::INFANTRY_4; - _robot_t[5] = RobotType::INFANTRY_5; - _robot_t[6] = RobotType::OUTPOST; - _robot_t[7] = RobotType::BASE; - _robot_t[8] = RobotType::SENTRY; + _ort = std::make_unique(model); + for (int i = 0; i < 9; ++i) + _robot_t[i] = static_cast(i); } /** diff --git a/extra/detector/src/armor_detector/detect.cpp b/extra/detector/src/armor_detector/detect.cpp index caace367..3229c819 100755 --- a/extra/detector/src/armor_detector/detect.cpp +++ b/extra/detector/src/armor_detector/detect.cpp @@ -29,7 +29,7 @@ DetectInfo ArmorDetector::detect(std::vector &groups, cv::Mat &src, // 二值化处理图像 PixChannel ch_minus = color == RED ? BLUE : RED; int thesh = color == RED ? para::armor_detector_param.GRAY_THRESHOLD_RED : para::armor_detector_param.GRAY_THRESHOLD_BLUE; - info.bin = rm::binary(src, color, ch_minus, thesh); + info.bin = binary(src, color, ch_minus, thesh); // 找到所有的灯条和装甲板 find(info.bin, info.features, info.combos, info.rois); diff --git a/extra/detector/src/armor_detector/find.cpp b/extra/detector/src/armor_detector/find.cpp index 6a054da2..0a145d3f 100755 --- a/extra/detector/src/armor_detector/find.cpp +++ b/extra/detector/src/armor_detector/find.cpp @@ -38,8 +38,11 @@ void ArmorDetector::find(cv::Mat &src, std::vector &features, std: for (const auto &armor : armors) { cv::Mat roi = Armor::getNumberROI(src, armor); - auto type = _ort->inference({roi}, {para::armor_detector_param.MODEL_MEAN}, {para::armor_detector_param.MODEL_STD})[0]; - armor->setType(_robot_t[type]); + PreprocessOptions preop; + preop.means = {para::armor_detector_param.MODEL_MEAN}; + preop.stds = {para::armor_detector_param.MODEL_STD}; + int idx = ClassificationNet::cast(_ort->inference({roi}, preop, {})).first; + armor->setType(_robot_t[idx]); rois.emplace_back(roi); } // eraseFakeArmors(armors); diff --git a/extra/detector/src/gyro_detector/find.cpp b/extra/detector/src/gyro_detector/find.cpp index 6e91f703..3d92739e 100755 --- a/extra/detector/src/gyro_detector/find.cpp +++ b/extra/detector/src/gyro_detector/find.cpp @@ -9,8 +9,6 @@ * */ -#include - #include #include "rmvl/detector/gyro_detector.h" @@ -40,8 +38,11 @@ void GyroDetector::find(cv::Mat &src, std::vector &features, std:: for (const auto &armor : armors) { cv::Mat roi = Armor::getNumberROI(src, armor); - auto type = _ort->inference({roi}, {para::gyro_detector_param.MODEL_MEAN}, {para::gyro_detector_param.MODEL_STD})[0]; - armor->setType(_robot_t[type]); + PreprocessOptions preop; + preop.means = {para::gyro_detector_param.MODEL_MEAN}; + preop.stds = {para::gyro_detector_param.MODEL_STD}; + int idx = ClassificationNet::cast(_ort->inference({roi}, preop, {})).first; + armor->setType(_robot_t[idx]); rois.emplace_back(roi); } // eraseFakeArmors(armors); diff --git a/extra/detector/src/gyro_detector/match.cpp b/extra/detector/src/gyro_detector/match.cpp index 9284a25c..871cdf6f 100755 --- a/extra/detector/src/gyro_detector/match.cpp +++ b/extra/detector/src/gyro_detector/match.cpp @@ -140,8 +140,8 @@ void GyroDetector::match(std::vector &groups, std::vector represent_set; - for (const auto &[represent, combo_set] : combo_maps) - represent_set.insert(represent); + for (const auto &p : combo_maps) + represent_set.insert(p.first); // represent // 距离最近的装甲板组匹配到相应的序列组中 for (auto &p_group : groups) { @@ -188,8 +188,8 @@ void GyroDetector::match(std::vector &groups, std::vector represent_set; - for (const auto &[represent, combo_set] : combo_maps) - represent_set.insert(represent); + for (const auto &p : combo_maps) + represent_set.insert(p.first); // represent size_t before_size = groups.size(); for (size_t i = 0; i < before_size; i++) { diff --git a/include/rmvl/rmvl.hpp b/include/rmvl/rmvl.hpp index 1f3c23a0..bd0ae2d0 100644 --- a/include/rmvl/rmvl.hpp +++ b/include/rmvl/rmvl.hpp @@ -30,10 +30,6 @@ #include "rmvl/light.hpp" #endif // HAVE_RMVL_LIGHT -#ifdef HAVE_RMVL_IMGPROC -#include "rmvl/imgproc.hpp" -#endif // HAVE_RMVL_IMGPROC - #ifdef HAVE_RMVL_OPCUA #include "rmvl/opcua.hpp" #endif // HAVE_RMVL_OPCUA diff --git a/modules/camera/include/rmvl/camera/camutils.hpp b/modules/camera/include/rmvl/camera/camutils.hpp index 864b8f66..8980078b 100644 --- a/modules/camera/include/rmvl/camera/camutils.hpp +++ b/modules/camera/include/rmvl/camera/camutils.hpp @@ -80,7 +80,7 @@ struct CameraConfig * @param[in] modes 配置模式参数包 */ template - inline void set(Args... modes) { [[maybe_unused]] int _[] = {(conf(modes), 0)...}; } + inline void set(Args... modes) { (..., conf(modes)); } private: inline void conf(TriggerChannel chn) { trigger_channel = chn; } diff --git a/modules/core/CMakeLists.txt b/modules/core/CMakeLists.txt index 30215d24..5cc8919a 100644 --- a/modules/core/CMakeLists.txt +++ b/modules/core/CMakeLists.txt @@ -16,7 +16,7 @@ rmvl_generate_para(core) rmvl_add_module( core EXTRA_HEADER ${EIGEN3_INCLUDE_DIR} - EXTERNAL ${CMAKE_THREAD_LIBS_INIT} opencv_core + EXTERNAL ${CMAKE_THREAD_LIBS_INIT} opencv_imgproc ) rmvl_compile_definitions( diff --git a/modules/imgproc/include/rmvl/imgproc/pretreat.h b/modules/core/include/rmvl/core/pretreat.hpp similarity index 73% rename from modules/imgproc/include/rmvl/imgproc/pretreat.h rename to modules/core/include/rmvl/core/pretreat.hpp index 5949c44c..0b311bed 100644 --- a/modules/imgproc/include/rmvl/imgproc/pretreat.h +++ b/modules/core/include/rmvl/core/pretreat.hpp @@ -1,11 +1,11 @@ /** - * @file pretreat.h - * @author RoboMaster Vision Community - * @brief Header of the image pretreating module + * @file pretreat.hpp + * @author zhaoxi (535394140@qq.com) + * @brief 图像预处理模块 * @version 1.0 - * @date 2022-11-23 + * @date 2024-06-05 * - * @copyright Copyright 2023 (c), RoboMaster Vision Community + * @copyright Copyright 2024 (c), zhaoxi * */ @@ -16,7 +16,15 @@ namespace rm { -//! @addtogroup imgproc +//! @addtogroup core +//! @{ +//! @defgroup core_pretreat 图像预处理模块 +//! @{ +//! @brief 提供了二值化等图像预处理功能 +//! @} core_pretreat +//! @} core + +//! @addtogroup core_pretreat //! @{ //! 像素通道枚举 @@ -52,6 +60,6 @@ cv::Mat binary(cv::Mat src, PixChannel ch1, PixChannel ch2, uint8_t threshold); */ cv::Mat binary(cv::Mat src, uint8_t threshold); -//! @} imgproc +//! @} core_pretreat } // namespace rm diff --git a/modules/imgproc/perf/perf_pretreat.cpp b/modules/core/perf/perf_pretreat.cpp similarity index 93% rename from modules/imgproc/perf/perf_pretreat.cpp rename to modules/core/perf/perf_pretreat.cpp index d2987ba9..46452c2d 100644 --- a/modules/imgproc/perf/perf_pretreat.cpp +++ b/modules/core/perf/perf_pretreat.cpp @@ -1,19 +1,19 @@ /** * @file perf_pretreat.cpp - * @author RoboMaster Vision Community - * @brief + * @author zhaoxi (535394140@qq.com) + * @brief * @version 1.0 - * @date 2022-11-24 - * - * @copyright Copyright 2023 (c), RoboMaster Vision Community - * + * @date 2024-06-05 + * + * @copyright Copyright 2024 (c), zhaoxi + * */ #include #include -#include "rmvl/imgproc/pretreat.h" +#include "rmvl/core/pretreat.hpp" namespace rm_test { diff --git a/modules/imgproc/src/pretreat.cpp b/modules/core/src/pretreat.cpp similarity index 97% rename from modules/imgproc/src/pretreat.cpp rename to modules/core/src/pretreat.cpp index 4b7b90a1..c18db54c 100644 --- a/modules/imgproc/src/pretreat.cpp +++ b/modules/core/src/pretreat.cpp @@ -12,7 +12,7 @@ #include #include "rmvl/core/util.hpp" -#include "rmvl/imgproc/pretreat.h" +#include "rmvl/core/pretreat.hpp" namespace rm { diff --git a/modules/core/test/test_optimal.cpp b/modules/core/test/test_optimal.cpp index 22f0ace3..3fcdaf6c 100644 --- a/modules/core/test/test_optimal.cpp +++ b/modules/core/test/test_optimal.cpp @@ -114,7 +114,7 @@ static inline double real_f(double x) { constexpr double FPS = 100; return 0.8 * std::sin(1.9 / FPS * x - 0.2) + 1.29; -}; +} TEST(Optimal, lsqnonlin_sine) { diff --git a/modules/imgproc/test/test_pretreat.cpp b/modules/core/test/test_pretreat.cpp similarity index 91% rename from modules/imgproc/test/test_pretreat.cpp rename to modules/core/test/test_pretreat.cpp index 271e90aa..38373901 100644 --- a/modules/imgproc/test/test_pretreat.cpp +++ b/modules/core/test/test_pretreat.cpp @@ -1,17 +1,17 @@ /** * @file test_pretreat.cpp - * @author RoboMaster Vision Community - * @brief + * @author zhaoxi (535394140@qq.com) + * @brief * @version 1.0 - * @date 2022-11-23 - * - * @copyright Copyright 2023 (c), RoboMaster Vision Community - * + * @date 2024-06-05 + * + * @copyright Copyright 2024 (c), zhaoxi + * */ #include -#include "rmvl/imgproc/pretreat.h" +#include "rmvl/core/pretreat.hpp" using namespace cv; using namespace rm; diff --git a/modules/imgproc/CMakeLists.txt b/modules/imgproc/CMakeLists.txt deleted file mode 100644 index cf1aeeb0..00000000 --- a/modules/imgproc/CMakeLists.txt +++ /dev/null @@ -1,23 +0,0 @@ -rmvl_add_module( - imgproc - DEPENDS core - EXTERNAL opencv_imgproc -) - -# Build the test program -if(BUILD_TESTS) - rmvl_add_test( - imgproc Unit - DEPENDS imgproc - DEPEND_TESTS GTest::gtest_main - ) -endif(BUILD_TESTS) - -# Build the perf test program -if(BUILD_PERF_TESTS) - rmvl_add_test( - imgproc Performance - DEPENDS imgproc - DEPEND_TESTS benchmark::benchmark_main - ) -endif(BUILD_PERF_TESTS) diff --git a/modules/imgproc/include/rmvl/imgproc.hpp b/modules/imgproc/include/rmvl/imgproc.hpp deleted file mode 100644 index 45b35f3a..00000000 --- a/modules/imgproc/include/rmvl/imgproc.hpp +++ /dev/null @@ -1,19 +0,0 @@ -/** - * @file imgproc.hpp - * @author RoboMaster Vision Community - * @brief - * @version 1.0 - * @date 2022-11-23 - * - * @copyright Copyright 2023 (c), RoboMaster Vision Community - * - */ - -//! @defgroup imgproc 图像处理 - -#include - -#ifdef HAVE_RMVL_IMGPROC -#include "imgproc/pretreat.h" -#include "imgproc/paint.hpp" -#endif // HAVE_RMVL_IMGPROC diff --git a/modules/imgproc/include/rmvl/imgproc/paint.hpp b/modules/imgproc/include/rmvl/imgproc/paint.hpp deleted file mode 100644 index 915972b1..00000000 --- a/modules/imgproc/include/rmvl/imgproc/paint.hpp +++ /dev/null @@ -1,175 +0,0 @@ -/** - * @file paint.hpp - * @author RoboMaster Vision Community - * @brief 绘画辅助工具 - * @version 1.0 - * @date 2023-01-17 - * - * @copyright Copyright 2023 (c), RoboMaster Vision Community - * - */ - -#pragma once - -#include - -#include - -namespace rm -{ - -//! @addtogroup imgproc -//! @{ - -//! 阵列工具 -class ArrayTool -{ -public: - /** - * @brief 平面线性阵列 - * - * @param[in] sketch 阵列特征 - * @param[in] spacing 阵列间距（Δx, Δy） - * @param[in] times 阵列个数（包括原特征） - * @return 阵列结果 - */ - template - static std::vector>> linear2D(const std::vector> &sketch, - const cv::Point_ &spacing, std::size_t times) - { - // 数据准备 - std::vector>> sketchs; - sketchs.reserve(times); - sketchs.push_back(sketch); - // 循环阵列 - for (std::size_t i = 1; i < times; ++i) - { - std::vector> tmp; - tmp.reserve(sketch.size()); - for (auto &point : sketch) - tmp.emplace_back(point.x + spacing.x * i, point.y + spacing.y * i); - sketchs.emplace_back(tmp); - } - return sketchs; - } - - /** - * @brief 平面圆周阵列 - * - * @param[in] sketch 阵列特征 - * @param[in] center 阵列中心 - * @param[in] spacing 阵列间距（Δθ，角度制，像素坐标系） - * @param[in] times 阵列个数（包括原特征） - * @return 阵列结果 - */ - template - static std::vector>> circular2D(const std::vector> &sketch, - const cv::Point_ ¢er, double spacing, std::size_t times) - { - // 数据准备 - std::vector>> sketchs; - sketchs.reserve(times); - sketchs.push_back(sketch); - // 循环阵列 - for (size_t i = 1; i < times; ++i) - { - // 计算旋转矩阵，左手系 - double spacing_rad = spacing * static_cast(i) * M_PI / 180.0; - double c_r = cos(spacing_rad); - double s_r = sin(spacing_rad); - cv::Matx22d R = {c_r, -s_r, - s_r, c_r}; - std::vector> tmp; - tmp.reserve(sketch.size()); - for (auto &point : sketch) - { - cv::Vec2d vec = {static_cast(point.x - center.x), - static_cast(point.y - center.y)}; - vec = R * vec; - tmp.emplace_back(static_cast(center.x + vec(0)), - static_cast(center.y + vec(1))); - } - sketchs.emplace_back(tmp); - } - return sketchs; - } - - /** - * @brief 空间线性阵列 - * - * @param[in] sketch 阵列特征 - * @param[in] spacing 阵列间距（Δx, Δy, Δz） - * @param[in] times 阵列个数（包括原特征） - * @return 阵列结果 - */ - template - static std::vector>> linear3D(const std::vector> &sketch, - const cv::Point3_ &spacing, std::size_t times) - { - std::vector>> sketchs; - sketchs.reserve(times); - sketchs.push_back(sketch); - for (std::size_t i = 1; i < times; ++i) - { - std::vector> tmp; - tmp.reserve(sketch.size()); - for (auto &point : sketch) - tmp.emplace_back(point.x + spacing.x * i, - point.y + spacing.y * i, - point.z + spacing.z * i); - sketchs.emplace_back(tmp); - } - return sketchs; - } - - /** - * @brief 空间圆周阵列 - * - * @param[in] sketch 阵列特征 - * @param[in] center 阵列中心 - * @param[in] axis 阵列转轴 - * @param[in] spacing 阵列间距（Δθ，角度制） - * @param[in] times 阵列个数（包括原特征） - * @return 阵列结果 - */ - template - static std::vector>> circular3D(const std::vector> &sketch, - const cv::Point3_ ¢er, const cv::Vec3d &axis, - double spacing, std::size_t times) - { - // 数据准备 - std::vector>> sketchs; - sketchs.reserve(times); - sketchs.push_back(sketch); - // 循环阵列 - for (size_t i = 1; i < times; ++i) - { - // 计算旋转矩阵 - double spacing_rad = spacing * static_cast(i) * M_PI / 180.0; - double c_r = cos(spacing_rad); - double s_r = sin(spacing_rad); - auto &a = axis; - cv::Matx33d R = {c_r + (1 - c_r) * a(0) * a(0), (1 - c_r) * a(0) * a(1) - s_r * a(2), (1 - c_r) * a(0) * a(2) + s_r * a(1), - (1 - c_r) * a(0) * a(1) + s_r * a(2), c_r + (1 - c_r) * a(1) * a(1), (1 - c_r) * a(1) * a(2) - s_r * a(0), - (1 - c_r) * a(0) * a(2) - s_r * a(1), (1 - c_r) * a(1) * a(2) + s_r * a(0), c_r + (1 - c_r) * a(2) * a(2)}; - std::vector> tmp; - tmp.reserve(sketch.size()); - for (auto &point : sketch) - { - cv::Vec3d vec = {static_cast(point.x - center.x), - static_cast(point.y - center.y), - static_cast(point.z - center.z)}; - vec = R * vec; - tmp.emplace_back(static_cast(center.x + vec(0)), - static_cast(center.y + vec(1)), - static_cast(center.z + vec(2))); - } - sketchs.emplace_back(tmp); - } - return sketchs; - } -}; - -//! @} imgproc - -} // namespace rm diff --git a/modules/imgproc/test/test_paint.cpp b/modules/imgproc/test/test_paint.cpp deleted file mode 100644 index acd33d77..00000000 --- a/modules/imgproc/test/test_paint.cpp +++ /dev/null @@ -1,65 +0,0 @@ -/** - * @file test_paint.cpp - * @author RoboMaster Vision Community - * @brief - * @version 1.0 - * @date 2023-01-18 - * - * @copyright Copyright 2023 (c), RoboMaster Vision Community - * - */ - -#include "rmvl/imgproc/paint.hpp" - -#include - -namespace rm_test -{ - -using rm::ArrayTool; - -TEST(ArrayToolTest, linear2D) -{ - std::vector pts = {{1, 1}, {1, 2}, {1, 3}}; - - auto ptss = ArrayTool::linear2D(pts, {3, 1}, 3); - EXPECT_EQ(ptss.size(), 3); - for (size_t i = 0; i < ptss.size(); ++i) - for (size_t j = 0; j < pts.size(); ++j) - EXPECT_EQ(ptss[i][j], (pts[j] + cv::Point(3 * i, 1 * i))); -} - -TEST(ArrayToolTest, circular2D) -{ - std::vector pts = {{20.f, 20.f}, {-10.f, -10.f}}; - auto ptss = ArrayTool::circular2D(pts, {0, 0}, 90.0, 2); - EXPECT_EQ(ptss.size(), 2); - EXPECT_EQ(ptss[1][0], cv::Point2f(-20, 20)); - EXPECT_EQ(ptss[1][1], cv::Point2f(10, -10)); -} - -TEST(ArrayToolTest, linear3D) -{ - std::vector pts = {{1, 1, 2}, - {1, 2, 4}, - {1, 3, 6}}; - - auto ptss = ArrayTool::linear3D(pts, {3, 1, 2}, 3); - EXPECT_EQ(ptss.size(), 3); - for (size_t i = 0; i < ptss.size(); ++i) - for (size_t j = 0; j < pts.size(); ++j) - EXPECT_EQ(ptss[i][j], (pts[j] + cv::Point3i(3 * i, 1 * i, 2 * i))); -} - -TEST(ArrayToolTest, circular3D) -{ - std::vector pt = {cv::Point3f(10, 20, 30)}; - auto pts1 = ArrayTool::circular3D(pt, {50, 50, 50}, {0, 0, 1}, 90.0, 2); - EXPECT_EQ(pts1.size(), 2); - EXPECT_EQ(pts1[1].front(), cv::Point3f(80, 10, 30)); - - auto pts2 = ArrayTool::circular3D(pt, {50, 50, 50}, {0, 1, 0}, 90.0, 2); - EXPECT_EQ(pts2[1].front(), cv::Point3f(30, 20, 90)); -} - -} // namespace rm_test diff --git a/modules/ml/include/rmvl/ml.hpp b/modules/ml/include/rmvl/ml.hpp index a779457c..845a044b 100644 --- a/modules/ml/include/rmvl/ml.hpp +++ b/modules/ml/include/rmvl/ml.hpp @@ -1,11 +1,11 @@ /** * @file ml.hpp - * @author RoboMaster Vision Community + * @author RMVL Community * @brief - * @version 1.0 - * @date 2023-05-20 + * @version 2.0 + * @date 2024-06-02 * - * @copyright Copyright 2023 (c), RoboMaster Vision Community + * @copyright Copyright 2024 (c), RMVL Community * */ @@ -14,7 +14,7 @@ /** * @defgroup ml 机器学习与深度学习支持库 * @{ - * @defgroup ml_ort ONNX-Runtime 分类网络部署库 + * @defgroup ml_ort onnxruntime 多态部署库 * @} */ diff --git a/modules/ml/include/rmvl/ml/ort.h b/modules/ml/include/rmvl/ml/ort.h index 98a23994..54a5bc76 100755 --- a/modules/ml/include/rmvl/ml/ort.h +++ b/modules/ml/include/rmvl/ml/ort.h @@ -1,17 +1,20 @@ /** * @file ort.h - * @author RoboMaster Vision Community + * @author RMVL Community * @brief the deployment library header file of the ONNXruntime (Ort) * @version 1.0 * @date 2022-02-04 * - * @copyright Copyright 2023 (c), RoboMaster Vision Community + * @copyright Copyright 2023 (c), RMVL Community * */ #pragma once -#include +#include +#include + +#include "rmvl/core/pretreat.hpp" namespace rm { @@ -19,33 +22,137 @@ namespace rm //! @addtogroup ml_ort //! @{ -//! ONNX-Runtime (Ort) 部署库 \cite onnx-rt -class OnnxRT +//! Ort 提供者 +enum class OrtProvider : uint8_t +{ + CPU, //!< 由 `CPU` 执行 + CUDA, //!< 由 `CUDA` 执行 + TensorRT, //!< 由 `TensorRT` 执行 + OpenVINO //!< 由 `OpenVINO` 执行 +}; + +//! 预处理选项 +struct PreprocessOptions { + std::vector means; //!< 均值 + std::vector stds; //!< 标准差 +}; + +//! 后处理选项 +struct PostprocessOptions +{ + PixChannel color{}; //!< 颜色通道 + std::vector thresh{}; //!< 阈值向量 +}; + +//! ONNX-Runtime (Ort) 部署库基类 \cite onnx-rt +class OnnxNet +{ +protected: + Ort::MemoryInfo _memory_info; //!< 内存分配信息 + Ort::Env _env; //!< 环境配置 + Ort::SessionOptions _session_options; //!< 会话选项 + std::unique_ptr _session; //!< 会话 +#if ORT_API_VERSION < 12 + std::vector _inames; //!< 输入名称 + std::vector _onames; //!< 输出名称 +#else + std::vector _inames; //!< 输入名称 + std::vector _onames; //!< 输出名称 +#endif + public: /** - * @brief 创建 OnnxRT 对象 + * @brief 创建 OnnxNet 对象 * * @param[in] model_path 模型路径，如果该路径不存在，则程序将因错误而退出 + * @param[in] prov Ort 提供者 + */ + OnnxNet(std::string_view model_path, OrtProvider prov); + + //! 打印环境信息 + static void printEnvInfo() noexcept; + //! 打印模型信息 + void printModelInfo() noexcept; + + /** + * @brief 推理 + * + * @param[in] images 所有输入图像 + * @param[in] preop 预处理选项，不同网络有不同的预处理选项 + * @param[in] postop 后处理选项，不同网络有不同的后处理选项 + * @return 使用 `std::any` 表示的所有推理结果，需要根据具体的网络进行解析 + * @note 可使用 `::cast` 函数对返回类型进行转换 + */ + std::any inference(const std::vector &images, const PreprocessOptions &preop, const PostprocessOptions &postop); + +private: + /** + * @brief 预处理 + * + * @param[in] images 所有输入图像 + * @param[in] preop 预处理选项，不同网络有不同的预处理选项 + * @return 模型的输入 Tensors */ - OnnxRT(std::string_view model_path); - ~OnnxRT(); + virtual std::vector preProcess(const std::vector &images, const PreprocessOptions &preop) = 0; - void printModelInfo(); + /** + * @brief 后处理 + * + * @param[in] output_tensors 模型的输出 Tensors + * @param[in] postop 后处理选项，不同网络有不同的后处理选项 + * @return 使用 `std::any` 表示的所有推理结果，需要根据具体的网络进行解析 + */ + virtual std::any postProcess(const std::vector &output_tensors, const PostprocessOptions &postop) = 0; +}; +/** + * @brief 分类网络推理类 + * @note 需满足 + * @note + * - 输入层为 `[1, c, h, w]`，其中 `c` 为输入图像的通道数，可以是 `1` 或者 `3`，`h` 为高度，`w` 为宽度 + * @note + * - 输出层为 `[1, n]`，其中 `n` 为类别数 + */ +class ClassificationNet : public OnnxNet +{ + std::vector> _iarrays; //!< 输入数组 + +public: /** - * @brief 预处理，推理和后处理 + * @brief 推理结果转换 * - * @param[in] images 所有的输入图像 - * @param[in] means 网络模型各通道的均值 - * @param[in] stds 网络模型各通道的标准差 - * @return 与概率最高的值对应的索引向量 + * @param[in] result 使用 `std::any` 表示的推理结果 + * @return 转换后的推理结果，为 `std::pair` 类型，表示分类结果及其置信度 */ - std::vector inference(const std::vector &images, const std::vector &means, const std::vector &stds); + static std::pair cast(const std::any &result) { return std::any_cast>(result); } + + /** + * @brief 创建分类网络对象 + * + * @param[in] model_path 模型路径，如果该路径不存在，则程序将因错误而退出 + * @param[in] prov Ort 提供者，默认为 `OrtProvider::CPU` + */ + ClassificationNet(std::string_view model_path, OrtProvider prov = OrtProvider::CPU); private: - class Impl; - Impl *_pimpl; + /** + * @brief 预处理 + * + * @param[in] images 所有输入图像 + * @param[in] options 预处理选项，包含各通道的均值和标准差 + * @return 模型的输入 Tensors + */ + std::vector preProcess(const std::vector &images, const PreprocessOptions &options) override; + + /** + * @brief 后处理 + * + * @param[in] output_tensors 模型的输出 Tensors + * @param[in] postop 无需后处理选项，传入 `{}` 即可 + * @return 用 `std::any` 表示的分类结果及其置信度，可使用 `rm::ClassificationNet::cast` 函数对返回类型进行转换 + */ + std::any postProcess(const std::vector &output_tensors, const PostprocessOptions &postop) override; }; //! @} ml_ort diff --git a/modules/ml/src/ort/classification.cpp b/modules/ml/src/ort/classification.cpp new file mode 100644 index 00000000..e750976b --- /dev/null +++ b/modules/ml/src/ort/classification.cpp @@ -0,0 +1,110 @@ +/** + * @file classification.cpp + * @author zhaoxi (535394140@qq.com) + * @brief + * @version 1.0 + * @date 2024-06-03 + * + * @copyright Copyright 2024 (c), zhaoxi + * + */ + +#include +#include + +#include "rmvl/core/util.hpp" +#include "rmvl/ml/ort.h" + +namespace rm +{ + +ClassificationNet::ClassificationNet(std::string_view model_path, OrtProvider prov) : OnnxNet(model_path, prov) +{ + // 初始化输入数组 + _iarrays.resize(_session->GetInputCount()); + for (size_t i = 0; i < _session->GetInputCount(); i++) + { + auto shape = _session->GetInputTypeInfo(i).GetTensorTypeAndShapeInfo().GetShape(); + _iarrays[i].resize(shape[1] * shape[2] * shape[3]); + } +} + +/** + * @brief 从 `NHWC` 格式的 `BGR` 图像生成 `NCHW` 格式的输入 `Tensor` + * + * @param[in] input_image 输入图像 + * @param[in] means 3 通道各自的均值 + * @param[in] stds 3 通道各自的标准差 + * @param[out] iarray `NCHW` 格式的输入 `Tensor` + */ +static void imageToVector(const cv::Mat &input_image, const std::vector &means, const std::vector &stds, std::vector &iarray) +{ + int H{input_image.rows}, W{input_image.cols}; + RMVL_DbgAssert(static_cast(3 * H * W) == iarray.size()); + RMVL_Assert(means.size() == 3 && stds.size() == 3); + // 转 Tensor 的 NCHW 格式，做归一化和标准化 + float *p_input_array = iarray.data(); + for (int c = 0; c < 3; c++) + for (int h = 0; h < H; h++) + for (int w = 0; w < W; w++) + p_input_array[c * H * W + h * W + w] = (input_image.at(h, w * 3 + 2 - c) / 255.f - means[c]) / stds[c]; +} + +/** + * @brief 从 `NHWC` 格式的灰度图像生成 `NCHW` 格式的输入 `Tensor` + * + * @param[in] input_image 输入图像 + * @param[in] mean 均值 + * @param[in] std 标准差 + * @param[out] iarray `NCHW` 格式的输入 `Tensor` + */ +static void imageToVector(const cv::Mat &input_image, float mean, float std, std::vector &iarray) +{ + int H{input_image.rows}, W{input_image.cols}; + RMVL_DbgAssert(static_cast(H * W) == iarray.size()); + // 转 Tensor 的 NCHW 格式，做归一化和标准化 + float *p_input_array = iarray.data(); + for (int h = 0; h < H; h++) + for (int w = 0; w < W; w++) + p_input_array[h * W + w] = (input_image.at(h, w) / 255.f - mean) / std; +} + +std::vector ClassificationNet::preProcess(const std::vector &images, const PreprocessOptions &options) +{ + std::size_t input_count = _session->GetInputCount(); + RMVL_Assert(input_count == 1 && images.size() == 1); + // 获取输入层 Tensor + std::vector input_tensors; + const cv::Mat &img = images.front(); + // 合法性检查 + auto shape = _session->GetInputTypeInfo(0).GetTensorTypeAndShapeInfo().GetShape(); + RMVL_Assert(shape.size() == 4); + if (shape[1] != img.channels() || shape[2] != img.rows || shape[3] != img.cols) + RMVL_Error_(RMVL_StsBadSize, "image (%d, %d, %d) unequal to shape (%d, %d, %d).", + img.channels(), img.rows, img.cols, shape[1], shape[2], shape[3]); + RMVL_Assert(shape[1] == 3 || shape[1] == 1); + shape[0] = 1; + // img -> iarray + RMVL_Assert(!options.means.empty() && !options.stds.empty()); + if (shape[1] == 3) + imageToVector(img, options.means, options.stds, _iarrays.front()); + else + imageToVector(img, options.means.front(), options.stds.front(), _iarrays.front()); + // 更新每个输入层的数据 + input_tensors.emplace_back(Ort::Value::CreateTensor( + _memory_info, _iarrays.front().data(), _iarrays.front().size(), shape.data(), shape.size())); + + return input_tensors; +} + +std::any ClassificationNet::postProcess(const std::vector &output_tensors, const PostprocessOptions &) +{ + RMVL_Assert(output_tensors.size() == 1); + auto &output_tensor = output_tensors.front(); + const float *output = output_tensor.GetTensorData(); + std::size_t size{output_tensor.GetTensorTypeAndShapeInfo().GetElementCount()}; + auto maxit = std::max_element(output, output + size); + return std::make_pair(maxit - output, *maxit); +} + +} // namespace rm diff --git a/modules/ml/src/ort/ort.cpp b/modules/ml/src/ort/ort.cpp index de04020a..cf9723a9 100755 --- a/modules/ml/src/ort/ort.cpp +++ b/modules/ml/src/ort/ort.cpp @@ -14,144 +14,132 @@ #include -#include "ort_impl.h" #include "rmvl/core/util.hpp" +#include "rmvl/ml/ort.h" -rm::OnnxRT::OnnxRT(std::string_view model_path) : _pimpl(new Impl(model_path)) {} -rm::OnnxRT::~OnnxRT() { delete _pimpl; } -void rm::OnnxRT::printModelInfo() { _pimpl->printModelInfo(); } -std::vector rm::OnnxRT::inference(const std::vector &images, const std::vector &means, - const std::vector &stds) +namespace rm { - try - { - return _pimpl->inference(images, means, stds); - } - catch (const rm::Exception &e) - { - delete _pimpl; - throw e; - } -} -rm::OnnxRT::Impl::Impl(std::string_view model_path) - : _env(OrtLoggingLevel::ORT_LOGGING_LEVEL_WARNING, "OnnxDeployment"), - _memory_info(Ort::MemoryInfo::CreateCpu(OrtAllocatorType::OrtArenaAllocator, OrtMemType::OrtMemTypeDefault)) -{ - if (model_path.empty()) - RMVL_Error(RMVL_StsBadArg, "Model path is empty!"); - setupEngine(model_path); -} +// 默认配置的内存分配器 +static Ort::AllocatorWithDefaultOptions alloc; -void rm::OnnxRT::Impl::setupEngine(std::string_view model_path) noexcept +OnnxNet::OnnxNet(std::string_view model_path, OrtProvider prov) : _memory_info(Ort::MemoryInfo::CreateCpu(OrtAllocatorType::OrtArenaAllocator, OrtMemType::OrtMemTypeDefault)) { -#ifdef WITH_ORT_CUDA - OrtSessionOptionsAppendExecutionProvider_CUDA(session_options, 0); -#endif // WITH_ORT_CUDA + if (model_path.empty()) + RMVL_Error(RMVL_StsBadArg, "Model path empty!"); -#ifdef WITH_ORT_TensorRT - OrtSessionOptionsAppendExecutionProvider_Tensorrt(session_options, 0); -#endif // WITH_ORT_TensorRT + switch (prov) + { + case OrtProvider::CUDA: + _session_options.AppendExecutionProvider_CUDA({}); + break; + case OrtProvider::TensorRT: + _session_options.AppendExecutionProvider_TensorRT({}); + break; + case OrtProvider::OpenVINO: { + OrtOpenVINOProviderOptions options; + options.device_type = "CPU_FP32"; + _session_options.AppendExecutionProvider_OpenVINO(options); + break; + } + default: + break; + } _session_options.SetGraphOptimizationLevel(GraphOptimizationLevel::ORT_ENABLE_ALL); - _p_session = std::make_unique(_env, model_path.data(), _session_options); + _session = std::make_unique(_env, model_path.data(), _session_options); - // define the names of the I/O nodes - for (size_t i = 0; i < _p_session->GetInputCount(); i++) - _input_names.emplace_back(_p_session->GetInputName(i, _allocator)); - for (size_t i = 0; i < _p_session->GetOutputCount(); i++) - _output_names.emplace_back(_p_session->GetOutputName(i, _allocator)); - // setup input array - _input_arrays.resize(_p_session->GetInputCount()); - for (size_t i = 0; i < _p_session->GetInputCount(); i++) - { - auto shape = _p_session->GetInputTypeInfo(i).GetTensorTypeAndShapeInfo().GetShape(); - _input_arrays[i].resize(shape[1] * shape[2] * shape[3]); - } + // 定义输入输出节点的名称 +#if ORT_API_VERSION < 12 + for (std::size_t i = 0; i < _session->GetInputCount(); i++) + _inames.emplace_back(_session->GetInputName(i, alloc)); + for (std::size_t i = 0; i < _session->GetOutputCount(); i++) + _onames.emplace_back(_session->GetOutputName(i, alloc)); +#else + for (std::size_t i = 0; i < _session->GetInputCount(); i++) + _inames.emplace_back(_session->GetInputNameAllocated(i, alloc)); + for (std::size_t i = 0; i < _session->GetOutputCount(); i++) + _onames.emplace_back(_session->GetOutputNameAllocated(i, alloc)); +#endif } -std::vector rm::OnnxRT::Impl::inference(const std::vector &images, - const std::vector &means, - const std::vector &stds) +std::any OnnxNet::inference(const std::vector &images, const PreprocessOptions &preop, const PostprocessOptions &postop) { - std::vector input_tensors = preProcess(images, means, stds); - std::vector output_tensors = doInference(input_tensors); - return postProcess(output_tensors); + auto itensors = preProcess(images, preop); +#if ORT_API_VERSION < 12 + return postProcess(_session->Run(Ort::RunOptions{nullptr}, _inames.data(), itensors.data(), itensors.size(), _onames.data(), _onames.size()), postop); +#else + std::vector input_names(_inames.size()); + for (std::size_t i = 0; i < _inames.size(); i++) + input_names[i] = _inames[i].get(); + std::vector output_names(_onames.size()); + for (std::size_t i = 0; i < _onames.size(); i++) + output_names[i] = _onames[i].get(); + return postProcess(_session->Run(Ort::RunOptions{nullptr}, input_names.data(), itensors.data(), itensors.size(), output_names.data(), output_names.size()), postop); +#endif } -std::vector rm::OnnxRT::Impl::preProcess(const std::vector &images, - const std::vector &means, - const std::vector &stds) +void OnnxNet::printEnvInfo() noexcept { - size_t input_count = _p_session->GetInputCount(); - if (input_count != images.size()) - RMVL_Error(RMVL_StsBadArg, "Size of the \"images\" are not equal to the model input_count."); - // get the correct data of each input layer - std::vector input_tensors; - for (size_t i = 0; i < input_count; i++) - { - auto input_shape = _p_session->GetInputTypeInfo(i).GetTensorTypeAndShapeInfo().GetShape(); - // [2], [3] are the correct size of the image - if (input_shape.size() != 4) - RMVL_Error(RMVL_StsBadSize, "Size of the input_shape of the model is not equal to \'4\'"); - if (input_shape[2] != input_shape[3]) - RMVL_Error(RMVL_StsError, "Shape of the input_shape[2] and input_shape[3] of the model is unequal"); - input_shape[0] = 1; - // update the size of each input layer - cv::Mat input_image; - resize(images[i], input_image, cv::Size(input_shape[2], input_shape[3])); - // allocate memory and normalization - imageToVector(input_image, _input_arrays[i], means, stds); - input_tensors.emplace_back(Ort::Value::CreateTensor(_memory_info, - _input_arrays[i].data(), _input_arrays[i].size(), - input_shape.data(), input_shape.size())); - } - return input_tensors; + printf("----------------- Build -----------------\n"); +#if ORT_API_VERSION < 15 + printf("version: 1.%d (< 1.15.0)\n", ORT_API_VERSION); +#else + printf("version: %s\n", Ort::GetVersionString().c_str()); +#endif + auto providers = Ort::GetAvailableProviders(); + printf("\n--------------- Providers ---------------\n"); + for (std::size_t i = 0; i < providers.size(); i++) + printf(" [%zu] %s\n", i, providers[i].c_str()); } -std::vector rm::OnnxRT::Impl::postProcess(const std::vector &output_tensors) noexcept +constexpr const char *element_data_type[] = { + "undefined", "float", "uint8", "int8", "uint16", "int16", + "int32", "int64", "string", "bool", "float16", "double", + "uint32", "uint64", "complex64", "complex128", "bfloat16"}; + +void OnnxNet::printModelInfo() noexcept { - // 所有输出对应的置信度最高的索引 - std::vector output_indexs; - for (auto &output_tensor : output_tensors) + if (_session == nullptr) { - // 获取置信度最高的索引 - const float *output = output_tensor.GetTensorData(); - std::vector indexs(output_tensor.GetTensorTypeAndShapeInfo().GetElementCount()); - iota(indexs.begin(), indexs.end(), 0); - auto it = max_element(indexs.begin(), indexs.end(), [&output](size_t lhs, size_t rhs) { - return output[lhs] < output[rhs]; - }); - output_indexs.emplace_back(*it); + printf("the model is empty.\n"); + return; + } + printf("-------------- Input Layer --------------\n"); + std::size_t input_node = _session->GetInputCount(); + printf("input node: n = %zu\n", input_node); + for (size_t i = 0; i < input_node; i++) + { +#if ORT_API_VERSION < 12 + printf("[%zu]\t┬ name: %s\n", i, _session->GetInputName(i, alloc)); +#else + printf("[%zu]\t┬ name: %s\n", i, _session->GetInputNameAllocated(i, alloc).get()); +#endif + auto input_dims = _session->GetInputTypeInfo(i).GetTensorTypeAndShapeInfo().GetShape(); + printf("\t│ dim: ["); + for (auto dim : input_dims) + printf("%ld, ", dim); + printf("\b\b]\n"); + printf("\t└ element type: %s\n", element_data_type[_session->GetInputTypeInfo(i).GetTensorTypeAndShapeInfo().GetElementType()]); } - return output_indexs; -} -void rm::OnnxRT::Impl::imageToVector(cv::Mat &input_image, std::vector &input_array, - const std::vector &means, const std::vector &stds) -{ - // CHW - int C = input_image.channels(); - if (C != 1 && C != 3) - RMVL_Error_(RMVL_StsBadArg, "Bad channel of the input argument: \"input_image\", chn = %d", C); - int H = input_image.rows; - int W = input_image.cols; - size_t pixels = C * H * W; - if (pixels != input_array.size()) - RMVL_Error(RMVL_StsBadArg, "The size of the arguments: \"input_image\" and \"input_array\" are not equal"); - if (static_cast(means.size()) != C || static_cast(stds.size()) != C) - RMVL_Error_(RMVL_StsBadArg, "Bad size of the input argument: the size of \"means\" or \"stds\" must be %d", C); - // 转 Tensor 的 NCHW 格式，做归一化和标准化 - float *p_input_array = input_array.data(); - for (int c = 0; c < C; c++) + printf("\n------------- Output Layer -------------\n"); + std::size_t output_node = _session->GetOutputCount(); + std::printf("output node: n = %zu\n", _session->GetOutputCount()); + for (std::size_t i = 0; i < output_node; i++) { - for (int h = 0; h < H; h++) - { - for (int w = 0; w < W; w++) - { - p_input_array[c * H * W + h * W + w] = input_image.ptr(h)[w * C + 2 - c] / 255.f; - p_input_array[c * H * W + h * W + w] = (p_input_array[c * H * W + h * W + w] - means[c]) / stds[c]; - } - } +#if ORT_API_VERSION < 12 + printf("[%zu]\t┬ name: %s\n", i, _session->GetInputName(i, alloc)); +#else + printf("[%zu]\t┬ name: %s\n", i, _session->GetOutputNameAllocated(i, alloc).get()); +#endif + auto output_dims = _session->GetOutputTypeInfo(i).GetTensorTypeAndShapeInfo().GetShape(); + printf("\t│ dim: ["); + for (auto dim : output_dims) + printf("%ld, ", dim); + printf("\b\b]\n"); + printf("\t└ element type: %s\n", element_data_type[_session->GetOutputTypeInfo(i).GetTensorTypeAndShapeInfo().GetElementType()]); } } + +} // namespace rm diff --git a/modules/ml/src/ort/ort_impl.h b/modules/ml/src/ort/ort_impl.h deleted file mode 100644 index 447165c7..00000000 --- a/modules/ml/src/ort/ort_impl.h +++ /dev/null @@ -1,66 +0,0 @@ -/** - * @file ort_impl.h - * @author zhaoxi (535394140@qq.com) - * @brief the deployment library of the ONNXruntime (Ort) - * @version 1.0 - * @date 2024-01-25 - * - * @copyright Copyright 2024 (c), zhaoxi - * - */ - -#pragma once - -#include - -#include "rmvl/ml/ort.h" - -namespace rm -{ - -//! ONNX-Runtime (Ort) 部署库 \cite onnx-rt -class OnnxRT::Impl -{ - using session_ptr = std::unique_ptr; - - Ort::Env _env; //!< 环境配置 - Ort::SessionOptions _session_options; //!< Session 配置 - Ort::MemoryInfo _memory_info; //!< Tensor 内存分配信息 - Ort::AllocatorWithDefaultOptions _allocator; //!< 默认配置的内存分配器 - session_ptr _p_session; - - std::vector> _input_arrays; //!< 输入数组 - std::vector _input_names; //!< 输入名 - std::vector _output_names; //!< 输出名 - -public: - Impl(std::string_view model_path); - ~Impl() = default; - - void printModelInfo() noexcept; - - std::vector inference(const std::vector &images, const std::vector &means, const std::vector &stds); - -private: - //! 初始化 Ort 引擎 - void setupEngine(std::string_view model_path) noexcept; - - //! 分配内存，将图像平展为 NCHW 格式的一维数组，同时将数组归一化 - void imageToVector(cv::Mat &input_image, std::vector &input_array, - const std::vector &means, const std::vector &stds); - - //! 预处理 - std::vector preProcess(const std::vector &images, const std::vector &means, const std::vector &stds); - - //! 后处理 - std::vector postProcess(const std::vector &output_tensors) noexcept; - - //! 推理并返回输出 Tensors - inline std::vector doInference(const std::vector &input_tensors) noexcept - { - return _p_session->Run(Ort::RunOptions{nullptr}, _input_names.data(), input_tensors.data(), - input_tensors.size(), _output_names.data(), _output_names.size()); - } -}; - -} // namespace rm diff --git a/modules/ml/src/ort/ort_print.cpp b/modules/ml/src/ort/ort_print.cpp deleted file mode 100755 index 28b7d151..00000000 --- a/modules/ml/src/ort/ort_print.cpp +++ /dev/null @@ -1,43 +0,0 @@ -/** - * @file ort_print.cpp - * @author RoboMaster Vision Community - * @brief - * @version 1.0 - * @date 2022-02-04 - * - * @copyright Copyright 2023 (c), RoboMaster Vision Community - * - */ - -#include "ort_impl.h" - -void rm::OnnxRT::Impl::printModelInfo() noexcept -{ - printf("-------------- Input Layer --------------\n"); - std::size_t input_node = _p_session->GetInputCount(); - printf("the number of input node is: %zu\n", input_node); - for (size_t i = 0; i < input_node; i++) - { - printf("[%zu]\t┬ name is: %s\n", i, _p_session->GetInputName(i, _allocator)); - auto input_dims = _p_session->GetInputTypeInfo(i).GetTensorTypeAndShapeInfo().GetShape(); - printf("\t│ dim is: [\n"); - for (auto dim : input_dims) - printf("%ld, ", dim); - printf("\b\b]\n"); - printf("\t└ type of each element is: %d\n", _p_session->GetInputTypeInfo(i).GetTensorTypeAndShapeInfo().GetElementType()); - } - - printf("\n------------- Output Layer -------------\n"); - std::size_t output_node = _p_session->GetOutputCount(); - std::printf("the number of output node is: %zu\n", _p_session->GetOutputCount()); - for (std::size_t i = 0; i < output_node; i++) - { - printf("[%zu]\t┬ name is: %s\n", i, _p_session->GetOutputName(i, _allocator)); - std::vector output_dims = _p_session->GetOutputTypeInfo(i).GetTensorTypeAndShapeInfo().GetShape(); - printf("\t│ dim is: ["); - for (auto dim : output_dims) - printf("%ld, ", dim); - printf("\b\b]\n"); - printf("\t└ type of each element is: %d\n", _p_session->GetOutputTypeInfo(i).GetTensorTypeAndShapeInfo().GetElementType()); - } -} diff --git a/modules/opcua/src/client.cpp b/modules/opcua/src/client.cpp index 6d3e2f61..a33bfaee 100644 --- a/modules/opcua/src/client.cpp +++ b/modules/opcua/src/client.cpp @@ -47,7 +47,7 @@ Client::Client(std::string_view address, UserConfig usr) } if (status != UA_STATUSCODE_GOOD) { - ERROR_("Failed to create client"); + ERROR_("Failed to create client: %s", UA_StatusCode_name(status)); UA_Client_delete(_client); _client = nullptr; } @@ -86,6 +86,8 @@ void Client::spinOnce() { UA_Client_run_iterate(_client, para::opcua_param.SPIN_ Variable Client::read(const NodeId &node) { + RMVL_DbgAssert(_client != nullptr); + UA_Variant p_val; UA_Variant_init(&p_val); UA_StatusCode status = UA_Client_readValueAttribute(_client, node, &p_val); @@ -98,6 +100,8 @@ Variable Client::read(const NodeId &node) bool Client::write(const NodeId &node, const Variable &val) { + RMVL_DbgAssert(_client != nullptr); + UA_Variant new_variant = helper::cvtVariable(val); auto status = UA_Client_writeValueAttribute(_client, node, &new_variant); UA_Variant_clear(&new_variant); @@ -111,6 +115,8 @@ bool Client::write(const NodeId &node, const Variable &val) bool Client::call(const NodeId &obj_node, const std::string &name, const std::vector &inputs, std::vector &outputs) { + RMVL_DbgAssert(_client != nullptr); + // 初始化输入、输出参数 std::vector input_variants; input_variants.reserve(inputs.size()); @@ -133,7 +139,7 @@ bool Client::call(const NodeId &obj_node, const std::string &name, const std::ve if (status != UA_STATUSCODE_GOOD) { ERROR_("Failed to call the method, node id: %d, error code: %s", - method_node.nid.identifier.numeric, UA_StatusCode_name(status)); + method_node.nid.identifier.numeric, UA_StatusCode_name(status)); return false; } outputs.reserve(output_size); @@ -146,6 +152,8 @@ bool Client::call(const NodeId &obj_node, const std::string &name, const std::ve NodeId Client::addViewNode(const View &view) { + RMVL_DbgAssert(_client != nullptr); + // 准备数据 NodeId retval; UA_ViewAttributes attr = UA_ViewAttributes_default; @@ -178,6 +186,8 @@ NodeId Client::addViewNode(const View &view) bool Client::monitor(NodeId node, UA_Client_DataChangeNotificationCallback on_change, uint32_t queue_size) { + RMVL_DbgAssert(_client != nullptr); + // 创建订阅 UA_CreateSubscriptionResponse resp; auto status = createSubscription(resp); @@ -201,6 +211,7 @@ bool Client::monitor(NodeId node, UA_Client_DataChangeNotificationCallback on_ch bool Client::monitor(NodeId node, const std::vector &names, UA_Client_EventNotificationCallback on_event) { + RMVL_DbgAssert(_client != nullptr); // 创建订阅 UA_CreateSubscriptionResponse sub_resp; if (!createSubscription(sub_resp)) @@ -266,8 +277,7 @@ bool Client::createSubscription(UA_CreateSubscriptionResponse &response) response = UA_Client_Subscriptions_create(_client, request, nullptr, nullptr, nullptr); if (response.responseHeader.serviceResult != UA_STATUSCODE_GOOD) { - ERROR_("Failed to create subscription, error: %s", - UA_StatusCode_name(response.responseHeader.serviceResult)); + ERROR_("Failed to create subscription, error: %s", UA_StatusCode_name(response.responseHeader.serviceResult)); return false; } return true; diff --git a/modules/opcua/src/helper.cpp b/modules/opcua/src/helper.cpp index fa7c6ad6..a9530833 100644 --- a/modules/opcua/src/helper.cpp +++ b/modules/opcua/src/helper.cpp @@ -52,6 +52,7 @@ NodeId operator|(NodeId origin, FindNodeInClient &&fnic) request.browsePaths = &browse_path; request.browsePathsSize = 1; + RMVL_Assert(p_client != nullptr); auto response = UA_Client_Service_translateBrowsePathsToNodeIds(p_client, request); if (response.responseHeader.serviceResult == UA_STATUSCODE_GOOD) if (response.resultsSize == 1 && response.results[0].targetsSize == 1) diff --git a/modules/opcua/src/server.cpp b/modules/opcua/src/server.cpp index 78c5c112..b510d979 100644 --- a/modules/opcua/src/server.cpp +++ b/modules/opcua/src/server.cpp @@ -31,7 +31,7 @@ Server::Server(uint16_t port, std::string_view name, const std::vectorlogger = UA_Log_Stdout_withLevel(UA_LOGLEVEL_ERROR); + config->logger = UA_Log_Stdout_withLevel(UA_LOGLEVEL_INFO); #else config->logging = UA_Log_Stdout_new(UA_LOGLEVEL_ERROR); #endif @@ -103,6 +103,8 @@ void Server::deleteServer() NodeId Server::addVariableTypeNode(const VariableType &vtype) { + RMVL_DbgAssert(_server != nullptr); + UA_VariableTypeAttributes attr = UA_VariableTypeAttributes_default; UA_Variant variant = helper::cvtVariable(vtype); // 设置属性 @@ -132,6 +134,8 @@ NodeId Server::addVariableTypeNode(const VariableType &vtype) NodeId Server::addVariableNode(const Variable &val, const NodeId &parent_id) { + RMVL_DbgAssert(_server != nullptr); + // 变量节点属性 `UA_VariableAttributes` UA_VariableAttributes attr = UA_VariableAttributes_default; UA_Variant variant = helper::cvtVariable(val); @@ -177,6 +181,8 @@ NodeId Server::addVariableNode(const Variable &val, const NodeId &parent_id) Variable Server::read(const NodeId &node) { + RMVL_DbgAssert(_server != nullptr); + UA_Variant p_val; UA_Variant_init(&p_val); auto status = UA_Server_readValue(_server, node, &p_val); @@ -189,6 +195,8 @@ Variable Server::read(const NodeId &node) bool Server::write(const NodeId &node, const Variable &val) { + RMVL_DbgAssert(_server != nullptr); + auto variant = helper::cvtVariable(val); auto status = UA_Server_writeValue(_server, node, variant); UA_Variant_clear(&variant); @@ -199,6 +207,8 @@ bool Server::write(const NodeId &node, const Variable &val) bool Server::addVariableNodeValueCallBack(NodeId id, ValueCallBackBeforeRead before_read, ValueCallBackAfterWrite after_write) { + RMVL_DbgAssert(_server != nullptr); + UA_ValueCallback callback{before_read, after_write}; auto status = UA_Server_setVariableNode_valueCallback(_server, id, callback); if (status != UA_STATUSCODE_GOOD) @@ -208,6 +218,8 @@ bool Server::addVariableNodeValueCallBack(NodeId id, ValueCallBackBeforeRead bef NodeId Server::addDataSourceVariableNode(const Variable &val, DataSourceRead on_read, DataSourceWrite on_write, NodeId parent_id) { + RMVL_DbgAssert(_server != nullptr); + // 变量节点属性 `UA_VariableAttributes` UA_VariableAttributes attr = UA_VariableAttributes_default; // 设置属性 @@ -245,6 +257,8 @@ NodeId Server::addDataSourceVariableNode(const Variable &val, DataSourceRead on_ NodeId Server::addMethodNode(const Method &method, const NodeId &parent_id) { + RMVL_DbgAssert(_server != nullptr); + UA_MethodAttributes attr = UA_MethodAttributes_default; attr.displayName = UA_LOCALIZEDTEXT(helper::en_US(), helper::to_char(method.display_name)); attr.description = UA_LOCALIZEDTEXT(helper::zh_CN(), helper::to_char(method.description)); @@ -282,11 +296,14 @@ NodeId Server::addMethodNode(const Method &method, const NodeId &parent_id) void Server::setMethodNodeCallBack(const NodeId &id, UA_MethodCallback on_method) { + RMVL_DbgAssert(_server != nullptr); UA_Server_setMethodNodeCallback(_server, id, on_method); } NodeId Server::addObjectTypeNode(const ObjectType &otype) { + RMVL_DbgAssert(_server != nullptr); + // 定义对象类型节点 UA_ObjectTypeAttributes attr = UA_ObjectTypeAttributes_default; attr.displayName = UA_LOCALIZEDTEXT(helper::en_US(), helper::to_char(otype.display_name)); @@ -346,6 +363,8 @@ NodeId Server::addObjectTypeNode(const ObjectType &otype) NodeId Server::addObjectNode(const Object &obj, NodeId parent_id) { + RMVL_DbgAssert(_server != nullptr); + UA_ObjectAttributes attr{UA_ObjectAttributes_default}; attr.displayName = UA_LOCALIZEDTEXT(helper::en_US(), helper::to_char(obj.display_name)); attr.description = UA_LOCALIZEDTEXT(helper::zh_CN(), helper::to_char(obj.description)); @@ -402,6 +421,8 @@ NodeId Server::addObjectNode(const Object &obj, NodeId parent_id) NodeId Server::addViewNode(const View &view) { + RMVL_DbgAssert(_server != nullptr); + // 准备数据 NodeId retval; UA_ViewAttributes attr = UA_ViewAttributes_default; @@ -434,6 +455,8 @@ NodeId Server::addViewNode(const View &view) NodeId Server::addEventTypeNode(const EventType &etype) { + RMVL_DbgAssert(_server != nullptr); + NodeId retval; UA_ObjectTypeAttributes attr = UA_ObjectTypeAttributes_default; attr.displayName = UA_LOCALIZEDTEXT(helper::en_US(), helper::to_char(etype.display_name)); @@ -482,6 +505,8 @@ NodeId Server::addEventTypeNode(const EventType &etype) bool Server::triggerEvent(const NodeId &node_id, const Event &event) { + RMVL_DbgAssert(_server != nullptr); + NodeId type_id = nodeBaseEventType | find(event.type()->browse_name); if (type_id.empty()) { diff --git a/modules/opcua/test/test_opcua_client.cpp b/modules/opcua/test/test_opcua_client.cpp index d5b10630..748c0893 100644 --- a/modules/opcua/test/test_opcua_client.cpp +++ b/modules/opcua/test/test_opcua_client.cpp @@ -50,7 +50,7 @@ void setSvr(rm::Server &svr) svr.addMethodNode(method); // 添加对象节点，包含字符串变量和乘法方法 svr.start(); - std::this_thread::sleep_for(std::chrono::milliseconds(20)); + std::this_thread::sleep_for(std::chrono::milliseconds(10)); } // 路径搜索 @@ -58,7 +58,7 @@ TEST(OPC_UA_ClientTest, read_variable) { rm::Server svr(5000); setSvr(svr); - rm::Client client("opc.tcp://localhost:5000"); + rm::Client client("opc.tcp://127.0.0.1:5000"); // 读取测试服务器上的变量值 auto id = rm::nodeObjectsFolder | client.find("array"); rm::Variable variable = client.read(id); @@ -75,7 +75,7 @@ TEST(OPC_UA_ClientTest, variable_IO) { rm::Server svr(5001); setSvr(svr); - rm::Client client("opc.tcp://localhost:5001"); + rm::Client client("opc.tcp://127.0.0.1:5001"); // 读取测试服务器上的变量值 auto id = rm::nodeObjectsFolder | client.find("single"); EXPECT_TRUE(client.write(id, 99)); @@ -92,7 +92,7 @@ TEST(OPC_UA_ClientTest, call) { rm::Server svr(5002); setSvr(svr); - rm::Client client("opc.tcp://localhost:5002"); + rm::Client client("opc.tcp://127.0.0.1:5002"); // 调用测试服务器上的方法 std::vector input = {1, 2}; std::vector output; @@ -116,7 +116,7 @@ TEST(OPC_UA_ClientTest, variable_monitor) { rm::Server svr(5003); setSvr(svr); - rm::Client client("opc.tcp://localhost:5003"); + rm::Client client("opc.tcp://127.0.0.1:5003"); // 订阅测试服务器上的变量 auto node_id = rm::nodeObjectsFolder | client.find("single"); EXPECT_TRUE(client.monitor(node_id, onChange, 5)); @@ -157,7 +157,7 @@ TEST(OPC_UA_ClientTest, event_monitor) etype.description = "测试事件类型"; etype.add("aaa", 3); svr.addEventTypeNode(etype); - rm::Client client("opc.tcp://localhost:5004"); + rm::Client client("opc.tcp://127.0.0.1:5004"); client.monitor(rm::nodeServer, {"SourceName", "aaa"}, onEvent); // 触发事件 rm::Event event(etype); diff --git a/samples/opcua/opcua_client.cpp b/samples/opcua/opcua_client.cpp index 5a38f053..e9a9d2c4 100644 --- a/samples/opcua/opcua_client.cpp +++ b/samples/opcua/opcua_client.cpp @@ -2,7 +2,7 @@ int main() { - rm::Client client("opc.tcp://localhost:4840"); + rm::Client client("opc.tcp://127.0.0.1:4840"); if (!client.ok()) return -1; auto position_id = rm::nodeObjectsFolder | client.find("position");