mmdeploy/csrc/preprocess/cpu/opencv_utils.h

152 lines
3.6 KiB
C
Raw Normal View History

Merge sdk (#251) * check in cmake * move backend_ops to csrc/backend_ops * check in preprocess, model, some codebase and their c-apis * check in CMakeLists.txt * check in parts of test_csrc * commit everything else * add readme * update core's BUILD_INTERFACE directory * skip codespell on third_party * update trt_net and ort_net's CMakeLists * ignore clion's build directory * check in pybind11 * add onnx.proto. Remove MMDeploy's dependency on ncnn's source code * export MMDeployTargets only when MMDEPLOY_BUILD_SDK is ON * remove useless message * target include directory is wrong * change target name from mmdeploy_ppl_net to mmdeploy_pplnn_net * skip install directory * update project's cmake * remove useless code * set CMAKE_BUILD_TYPE to Release by force if it isn't set by user * update custom ops CMakeLists * pass object target's source lists * fix lint end-of-file * fix lint: trailing whitespace * fix codespell hook * remove bicubic_interpolate to csrc/backend_ops/ * set MMDEPLOY_BUILD_SDK OFF * change custom ops build command * add spdlog installation command * update docs on how to checkout pybind11 * move bicubic_interpolate to backend_ops/tensorrt directory * remove useless code * correct cmake * fix typo * fix typo * fix install directory * correct sdk's readme * set cub dir when cuda version < 11.0 * change directory where clang-format will apply to * fix build command * add .clang-format * change clang-format style from google to file * reformat csrc/backend_ops * format sdk's code * turn off clang-format for some files * add -Xcompiler=-fno-gnu-unique * fix trt topk initialize * check in config for sdk demo * update cmake script and csrc's readme * correct config's path * add cuda include directory, otherwise compile failed in case of tensorrt8.2 * clang-format onnx2ncnn.cpp Co-authored-by: zhangli <lzhang329@gmail.com> Co-authored-by: grimoire <yaoqian@sensetime.com>
2021-12-07 10:57:55 +08:00
// Copyright (c) OpenMMLab. All rights reserved.
#ifndef MMDEPLOY_OPENCV_UTILS_H
#define MMDEPLOY_OPENCV_UTILS_H
#include "core/mat.h"
#include "core/mpl/type_traits.h"
#include "core/serialization.h"
#include "core/tensor.h"
#include "opencv2/opencv.hpp"
namespace mmdeploy {
namespace cpu {
cv::Mat Mat2CVMat(const Mat& mat);
cv::Mat Tensor2CVMat(const Tensor& tensor);
Mat CVMat2Mat(const cv::Mat& mat, PixelFormat format);
Tensor CVMat2Tensor(const cv::Mat& mat);
/**
* @brief resize an image to specified size
*
* @param src input image
* @param dst_height output image's height
* @param dst_width output image's width
* @return output image if success, error code otherwise
*/
cv::Mat Resize(const cv::Mat& src, int dst_height, int dst_width, const std::string& interpolation);
/**
* @brief crop an image
*
* @param src input image
* @param top
* @param left
* @param bottom
* @param right
* @return cv::Mat
*/
cv::Mat Crop(const cv::Mat& src, int top, int left, int bottom, int right);
/**
* @brief Do normalization to an image
*
* @param src input image. It is assumed to be BGR if the channel is 3
* @param mean
* @param std
* @param to_rgb
* @param inplace
* @return cv::Mat
*/
cv::Mat Normalize(cv::Mat& src, const std::vector<float>& mean, const std::vector<float>& std,
bool to_rgb, bool inplace = true);
/**
* @brief tranpose an image, from {h, w, c} to {c, h, w}
*
* @param src input image
* @return
*/
cv::Mat Transpose(const cv::Mat& src);
/**
* @brief convert an image to another color space
*
* @param src
* @param src_format
* @param dst_format
* @return
*/
cv::Mat ColorTransfer(const cv::Mat& src, PixelFormat src_format, PixelFormat dst_format);
/**
*
* @param src
* @param top
* @param left
* @param bottom
* @param right
* @param border_type
* @param val
* @return
*/
cv::Mat Pad(const cv::Mat& src, int top, int left, int bottom, int right, int border_type,
float val);
/**
* @brief compare two images
*
* @param src1 one input image
* @param src2 the other input image
* @return bool true means the images are the same
*/
bool Compare(const cv::Mat& src1, const cv::Mat& src2);
} // namespace cpu
namespace detail {
template <typename T>
struct IsCvPoint : std::false_type {};
template <typename T>
struct IsCvPoint<::cv::Point_<T>> : std::true_type {};
} // namespace detail
template <typename Archive, typename T,
std::enable_if_t<detail::IsCvPoint<uncvref_t<T>>::value, int> = 0>
void serialize(Archive&& archive, T&& p) {
int size{2};
std::forward<Archive>(archive).init(size);
std::forward<Archive>(archive).item(std::forward<T>(p).x);
std::forward<Archive>(archive).item(std::forward<T>(p).y);
}
template <typename Archive, typename T, std::enable_if_t<detail::IsCvPoint<T>::value, int> = 0>
void save(Archive& archive, std::vector<T>& v) {
archive.init(array_tag<T>{v.size() * 2});
for (const auto& p : v) {
archive.item(p.x);
archive.item(p.y);
}
}
template <typename Archive, typename T, std::enable_if_t<detail::IsCvPoint<T>::value, int> = 0>
void save(Archive& archive, const std::vector<T>& v) {
archive.init(array_tag<T>{v.size() * 2});
for (const auto& p : v) {
archive.item(p.x);
archive.item(p.y);
}
}
template <typename Archive, typename T, std::enable_if_t<detail::IsCvPoint<T>::value, int> = 0>
void load(Archive& archive, std::vector<T>& v) {
size_t size{};
archive.init(size);
size /= 2;
T p;
for (int i = 0; i < size; ++i) {
archive.item(p.x);
archive.item(p.y);
v.push_back(p);
}
}
} // namespace mmdeploy
#endif // MMDEPLOY_OPENCV_UTILS_H