1、使用cuda加速
#include <fstream>
#include <sstream>
#include <iostream>
#include <opencv2/imgproc.hpp>
#include <opencv2/highgui.hpp>
#include <tensorrt_provider_factory.h> // 提供cuda加速
#include <onnxruntime_cxx_api.h> // C或c++的api
// 命名空间
using namespace std;
using namespace cv;
using namespace Ort;
// 自定义配置结构
struct Configuration
{
public:
float confThreshold; // Confidence threshold
float nmsThreshold; // Non-maximum suppression threshold
float objThreshold; //Object Confidence threshold
string modelpath;
};
// 定义BoxInfo结构类型
typedef struct BoxInfo
{
float x1;
float y1;
float x2;
float y2;
float score;
int label;
} BoxInfo;
// int endsWith(string s, string sub) {
// return s.rfind(sub) == (s.length() - sub.length()) ? 1 : 0;
// }
// const float anchors_640[3][6] = { {10.0, 13.0, 16.0, 30.0, 33.0, 23.0},
// {30.0, 61.0, 62.0, 45.0, 59.0, 119.0},
// {116.0, 90.0, 156.0, 198.0, 373.0, 326.0} };
// const float anchors_1280[4][6] = { {19, 27, 44, 40, 38, 94},{96, 68, 86, 152, 180, 137},{140, 301, 303, 264, 238, 542},
// {436, 615, 739, 380, 925, 792} };
class YOLOv5
{
public:
YOLOv5(Configuration config);
void detect(Mat& frame);
private:
float confThreshold;
float nmsThreshold;
float objThreshold;
int inpWidth;
int inpHeight;
int nout;
int num_proposal;
int num_classes;
string classes[80] = {"person", "bicycle", "car", "motorbike", "aeroplane", "bus",
"train", "truck", "boat", "traffic light", "fire hydrant",
"stop sign", "parking meter", "bench", "bird", "cat", "dog",
"horse", "sheep", "cow", "elephant", "bear", "zebra", "giraffe",
"backpack", "umbrella", "handbag", "tie", "suitcase", "frisbee",
"skis", "snowboard", "sports ball", "kite", "baseball bat",
"baseball glove", "skateboard", "surfboard", "tennis racket",
"bottle", "wine glass", "cup", "fork", "knife", "spoon", "bowl",
"banana", "apple", "sandwich", "orange", "broccoli", "carrot",
"hot dog", "pizza", "donut", "cake", "chair", "sofa", "pottedplant",
"bed", "diningtable", "toilet", "tvmonitor", "laptop", "mouse",
"remote", "keyboard", "cell phone", "microwave", "oven", "toaster",
"sink", "refrigerator", "book", "clock", "vase", "scissors",
"teddy bear", "hair drier", "toothbrush"};
const bool keep_ratio = true;
vector<float> input_image_; // 输入图片
void normalize_(Mat img); // 归一化函数
void nms(vector<BoxInfo>& input_boxes);
Mat resize_image(Mat srcimg, int *newh, int *neww, int *top, int *left);
Env env = Env(ORT_LOGGING_LEVEL_ERROR, "yolov5-6.1"); // 初始化环境
Session *ort_session = nullptr; // 初始化Session指针选项
SessionOptions sessionOptions = SessionOptions(); //初始化Session对象
//SessionOptions sessionOptions;
vector<char*> input_names; // 定义一个字符指针vector
vector<char*> output_names; // 定义一个字符指针vector
vector<vector<int64_t>> input_node_dims; // >=1 outputs ，二维vector
vector<vector<int64_t>> output_node_dims; // >=1 outputs ,int64_t C/C++标准
};
YOLOv5::YOLOv5(Configuration config)
{
this->confThreshold = config.confThreshold;
this->nmsThreshold = config.nmsThreshold;
this->objThreshold = config.objThreshold;
this->num_classes = sizeof(this->classes)/sizeof(this->classes[0]); // 类别数量
this->inpHeight = 640;
this->inpWidth = 640;
string model_path = config.modelpath;
//std::wstring widestr = std::wstring(model_path.begin(), model_path.end()); //用于UTF-16编码的字符
//gpu, https://blog.csdn.net/weixin_44684139/article/details/123504222
//CUDA加速开启
//OrtSessionOptionsAppendExecutionProvider_CUDA(sessionOptions, 0);
std::vector<std::string> availableProviders = Ort::GetAvailableProviders();
std::cout << "availableProviders = " << availableProviders[0] << std::endl;
//auto cudaAvailable = std::find(availableProviders.begin(), availableProviders.end(), "TensorrtExecutionProvider");
OrtCUDAProviderOptions cudaOption;
sessionOptions.AppendExecutionProvider_CUDA(cudaOption);
OrtSessionOptionsAppendExecutionProvider_CUDA(sessionOptions, 0);
sessionOptions.SetGraphOptimizationLevel(ORT_ENABLE_BASIC); //设置图优化类型
//ort_session = new Session(env, widestr.c_str(), sessionOptions); // 创建会话，把模型加载到内存中
//ort_session = new Session(env, (const ORTCHAR_T*)model_path.c_str(), sessionOptions); // 创建会话，把模型加载到内存中
ort_session = new Session(env, (const char*)model_path.c_str(), sessionOptions);
size_t numInputNodes = ort_session->GetInputCount(); //输入输出节点数量
size_t numOutputNodes = ort_session->GetOutputCount();
AllocatorWithDefaultOptions allocator; // 配置输入输出节点内存
for (int i = 0; i < numInputNodes; i++)
{
input_names.push_back(ort_session->GetInputName(i, allocator)); // 内存
//Ort::AllocatedStringPtr input_name_Ptr = ort_session->GetInputNameAllocated(i, allocator);
//input_names.push_back(input_name_Ptr.get());
Ort::TypeInfo input_type_info = ort_session->GetInputTypeInfo(i); // 类型
auto input_tensor_info = input_type_info.GetTensorTypeAndShapeInfo(); //
auto input_dims = input_tensor_info.GetShape(); // 输入shape
input_node_dims.push_back(input_dims); // 保存
}
for (int i = 0; i < numOutputNodes; i++)
{
output_names.push_back(ort_session->GetOutputName(i, allocator));
//Ort::AllocatedStringPtr output_name_Ptr = ort_session->GetOutputNameAllocated(i, allocator);
//output_names.push_back(output_name_Ptr.get());
Ort::TypeInfo output_type_info = ort_session->GetOutputTypeInfo(i);
auto output_tensor_info = output_type_info.GetTensorTypeAndShapeInfo();
auto output_dims = output_tensor_info.GetShape();
output_node_dims.push_back(output_dims);
}
this->inpHeight = input_node_dims[0][2];
this->inpWidth = input_node_dims[0][3];
this->nout = output_node_dims[0][2]; // 5+classes
this->num_proposal = output_node_dims[0][1]; // pre_box
}
Mat YOLOv5::resize_image(Mat srcimg, int *newh, int *neww, int *top, int *left)
{
int srch = srcimg.rows, srcw = srcimg.cols;
*newh = this->inpHeight;
*neww = this->inpWidth;
Mat dstimg;
if (this->keep_ratio && srch != srcw) {
float hw_scale = (float)srch / srcw;
if (hw_scale > 1) {
*newh = this->inpHeight;
*neww = int(this->inpWidth / hw_scale);
resize(srcimg, dstimg, Size(*neww, *newh), INTER_AREA);
*left = int((this->inpWidth - *neww) * 0.5);
copyMakeBorder(dstimg, dstimg, 0, 0, *left, this->inpWidth - *neww - *left, BORDER_CONSTANT, 114);
}
else {
*newh = (int)this->inpHeight * hw_scale;
*neww = this->inpWidth;
resize(srcimg, dstimg, Size(*neww, *newh), INTER_AREA);
*top = (int)(this->inpHeight - *newh) * 0.5;
copyMakeBorder(dstimg, dstimg, *top, this->inpHeight - *newh - *top, 0, 0, BORDER_CONSTANT, 114);
}
}
else {
resize(srcimg, dstimg, Size(*neww, *newh), INTER_AREA);
}
return dstimg;
}
void YOLOv5::normalize_(Mat img)
{
// img.convertTo(img, CV_32F);
int row = img.rows;
int col = img.cols;
this->input_image_.resize(row * col * img.channels()); // vector大小
for (int c = 0; c < 3; c++) // bgr
{
for (int i = 0; i < row; i++) // 行
{
for (int j = 0; j < col; j++) // 列
{
float pix = img.ptr<uchar>(i)[j * 3 + 2 - c]; // Mat里的ptr函数访问任意一行像素的首地址,2-c:表示rgb
this->input_image_[c * row * col + i * col + j] = pix / 255.0;
}
}
}
}
void YOLOv5::nms(vector<BoxInfo>& input_boxes)
{
sort(input_boxes.begin(), input_boxes.end(), [](BoxInfo a, BoxInfo b) { return a.score > b.score; }); // 降序排列
vector<float> vArea(input_boxes.size());
for (int i = 0; i < input_boxes.size(); ++i)
{
vArea[i] = (input_boxes[i].x2 - input_boxes[i].x1 + 1)
* (input_boxes[i].y2 - input_boxes[i].y1 + 1);
}
// 全初始化为false，用来作为记录是否保留相应索引下pre_box的标志vector
vector<bool> isSuppressed(input_boxes.size(), false);
for (int i = 0; i < input_boxes.size(); ++i)
{
if (isSuppressed[i]) { continue; }
for (int j = i + 1; j < input_boxes.size(); ++j)
{
if (isSuppressed[j]) { continue; }
float xx1 = max(input_boxes[i].x1, input_boxes[j].x1);
float yy1 = max(input_boxes[i].y1, input_boxes[j].y1);
float xx2 = min(input_boxes[i].x2, input_boxes[j].x2);
float yy2 = min(input_boxes[i].y2, input_boxes[j].y2);
float w = max(0.0f, xx2 - xx1 + 1);
float h = max(0.0f, yy2 - yy1 + 1);
float inter = w * h; // 交集
if(input_boxes[i].label == input_boxes[j].label)
{
float ovr = inter / (vArea[i] + vArea[j] - inter); // 计算iou
if (ovr >= this->nmsThreshold)
{
isSuppressed[j] = true;
}
}
}
}
// return post_nms;
int idx_t = 0;
// remove_if()函数 remove_if(beg, end, op) //移除区间[beg,end)中每一个“令判断式:op(elem)获得true”的元素
input_boxes.erase(remove_if(input_boxes.begin(), input_boxes.end(), [&idx_t, &isSuppressed](const BoxInfo& f) { return isSuppressed[idx_t++]; }), input_boxes.end());
// 另一种写法
// sort(input_boxes.begin(), input_boxes.end(), [](BoxInfo a, BoxInfo b) { return a.score > b.score; }); // 降序排列
// vector<bool> remove_flags(input_boxes.size(),false);
// auto iou = [](const BoxInfo& box1,const BoxInfo& box2)
// {
// float xx1 = max(box1.x1, box2.x1);
// float yy1 = max(box1.y1, box2.y1);
// float xx2 = min(box1.x2, box2.x2);
// float yy2 = min(box1.y2, box2.y2);
// // 交集
// float w = max(0.0f, xx2 - xx1 + 1);
// float h = max(0.0f, yy2 - yy1 + 1);
// float inter_area = w * h;
// // 并集
// float union_area = max(0.0f,box1.x2-box1.x1) * max(0.0f,box1.y2-box1.y1)
// + max(0.0f,box2.x2-box2.x1) * max(0.0f,box2.y2-box2.y1) - inter_area;
// return inter_area / union_area;
// };
// for (int i = 0; i < input_boxes.size(); ++i)
// {
// if(remove_flags[i]) continue;
// for (int j = i + 1; j < input_boxes.size(); ++j)
// {
// if(remove_flags[j]) continue;
// if(input_boxes[i].label == input_boxes[j].label && iou(input_boxes[i],input_boxes[j])>=this->nmsThreshold)
// {
// remove_flags[j] = true;
// }
// }
// }
// int idx_t = 0;
// // remove_if()函数 remove_if(beg, end, op) //移除区间[beg,end)中每一个“令判断式:op(elem)获得true”的元素
// input_boxes.erase(remove_if(input_boxes.begin(), input_boxes.end(), [&idx_t, &remove_flags](const BoxInfo& f) { return remove_flags[idx_t++]; }), input_boxes.end());
}
void YOLOv5::detect(Mat& frame)
{
int newh = 0, neww = 0, padh = 0, padw = 0;
Mat dstimg = this->resize_image(frame, &newh, &neww, &padh, &padw);
this->normalize_(dstimg);
// 定义一个输入矩阵，int64_t是下面作为输入参数时的类型
array<int64_t, 4> input_shape_{ 1, 3, this->inpHeight, this->inpWidth };
//创建输入tensor
auto allocator_info = MemoryInfo::CreateCpu(OrtDeviceAllocator, OrtMemTypeCPU);
Value input_tensor_ = Value::CreateTensor<float>(allocator_info, input_image_.data(), input_image_.size(), input_shape_.data(), input_shape_.size());
// 开始推理
vector<Value> ort_outputs = ort_session->Run(RunOptions{ nullptr }, &input_names[0], &input_tensor_, 1, output_names.data(), output_names.size()); // 开始推理
/generate proposals
vector<BoxInfo> generate_boxes; // BoxInfo自定义的结构体
float ratioh = (float)frame.rows / newh, ratiow = (float)frame.cols / neww;
float* pdata = ort_outputs[0].GetTensorMutableData<float>(); // GetTensorMutableData
for(int i = 0; i < num_proposal; ++i) // 遍历所有的num_pre_boxes
{
int index = i * nout; // prob[b*num_pred_boxes*(classes+5)]
float obj_conf = pdata[index + 4]; // 置信度分数
if (obj_conf > this->objThreshold) // 大于阈值
{
int class_idx = 0;
float max_class_socre = 0;
for (int k = 0; k < this->num_classes; ++k)
{
if (pdata[k + index + 5] > max_class_socre)
{
max_class_socre = pdata[k + index + 5];
class_idx = k;
}
}
max_class_socre *= obj_conf; // 最大的类别分数*置信度
if (max_class_socre > this->confThreshold) // 再次筛选
{
//const int class_idx = classIdPoint.x;
float cx = pdata[index]; //x
float cy = pdata[index+1]; //y
float w = pdata[index+2]; //w
float h = pdata[index+3]; //h
float xmin = (cx - padw - 0.5 * w)*ratiow;
float ymin = (cy - padh - 0.5 * h)*ratioh;
float xmax = (cx - padw + 0.5 * w)*ratiow;
float ymax = (cy - padh + 0.5 * h)*ratioh;
generate_boxes.push_back(BoxInfo{ xmin, ymin, xmax, ymax, max_class_socre, class_idx });
}
}
}
// Perform non maximum suppression to eliminate redundant overlapping boxes with
// lower confidences
nms(generate_boxes);
for (size_t i = 0; i < generate_boxes.size(); ++i)
{
int xmin = int(generate_boxes[i].x1);
int ymin = int(generate_boxes[i].y1);
rectangle(frame, Point(xmin, ymin), Point(int(generate_boxes[i].x2), int(generate_boxes[i].y2)), Scalar(0, 0, 255), 2);
string label = format("%.2f", generate_boxes[i].score);
label = this->classes[generate_boxes[i].label] + ":" + label;
putText(frame, label, Point(xmin, ymin - 5), FONT_HERSHEY_SIMPLEX, 0.75, Scalar(0, 255, 0), 1);
}
}
int main(int argc,char *argv[])
{
clock_t startTime,endTime; //计算时间
Configuration yolo_nets = { 0.3, 0.5, 0.3,"/home/jingjia/code/ken/yolov5-onnxruntime-master/models/yolov5s_int8q_jingtai.onnx" };
YOLOv5 yolo_model(yolo_nets);
string imgpath = "/home/jingjia/code/ken/yolov5-onnxruntime-master/images/bus.jpg";
string videoPath = "/home/jingjia/code/ken/yolov5-onnxruntime-master/images/1.mp4";
//Mat srcimg = imread(imgpath);
cv::VideoCapture cap;
cap.open(videoPath);
//cap>>frame;
Mat srcimg;
while(true){
cap>>srcimg;
double timeStart = (double)getTickCount();
startTime = clock();//计时开始
yolo_model.detect(srcimg);
endTime = clock();//计时结束
imshow("result", srcimg);
waitKey(1);
double nTime = ((double)getTickCount() - timeStart) / getTickFrequency();
cout << "clock_running time is:" <<(double)(endTime - startTime) / CLOCKS_PER_SEC << "s" << endl;
//cout << "The run time is:" << (double)clock() /CLOCKS_PER_SEC<< "s" << endl;
//cout << "getTickCount_running time :" << nTime << "sec\n" << endl;
}
// static const string kWinName = "Deep learning object detection in ONNXRuntime";
// namedWindow(kWinName, WINDOW_NORMAL);
// imshow(kWinName, srcimg);
//imwrite("restult_ort.jpg",srcimg);
// waitKey(0);
// destroyAllWindows();
return 0;
}?2、使用tensorrt加速
#include <fstream>
#include <sstream>
#include <iostream>
#include <opencv2/imgproc.hpp>
#include <opencv2/highgui.hpp>
#include <tensorrt_provider_factory.h> // 提供cuda加速
#include <onnxruntime_cxx_api.h> // C或c++的api
// 命名空间
using namespace std;
using namespace cv;
using namespace Ort;
// 自定义配置结构
struct Configuration
{
public:
float confThreshold; // Confidence threshold
float nmsThreshold; // Non-maximum suppression threshold
float objThreshold; //Object Confidence threshold
string modelpath;
};
// 定义BoxInfo结构类型
typedef struct BoxInfo
{
float x1;
float y1;
float x2;
float y2;
float score;
int label;
} BoxInfo;
// int endsWith(string s, string sub) {
// return s.rfind(sub) == (s.length() - sub.length()) ? 1 : 0;
// }
// const float anchors_640[3][6] = { {10.0, 13.0, 16.0, 30.0, 33.0, 23.0},
// {30.0, 61.0, 62.0, 45.0, 59.0, 119.0},
// {116.0, 90.0, 156.0, 198.0, 373.0, 326.0} };
// const float anchors_1280[4][6] = { {19, 27, 44, 40, 38, 94},{96, 68, 86, 152, 180, 137},{140, 301, 303, 264, 238, 542},
// {436, 615, 739, 380, 925, 792} };
class YOLOv5
{
public:
YOLOv5(Configuration config);
void detect(Mat& frame);
private:
float confThreshold;
float nmsThreshold;
float objThreshold;
int inpWidth;
int inpHeight;
int nout;
int num_proposal;
int num_classes;
string classes[80] = {"person", "bicycle", "car", "motorbike", "aeroplane", "bus",
"train", "truck", "boat", "traffic light", "fire hydrant",
"stop sign", "parking meter", "bench", "bird", "cat", "dog",
"horse", "sheep", "cow", "elephant", "bear", "zebra", "giraffe",
"backpack", "umbrella", "handbag", "tie", "suitcase", "frisbee",
"skis", "snowboard", "sports ball", "kite", "baseball bat",
"baseball glove", "skateboard", "surfboard", "tennis racket",
"bottle", "wine glass", "cup", "fork", "knife", "spoon", "bowl",
"banana", "apple", "sandwich", "orange", "broccoli", "carrot",
"hot dog", "pizza", "donut", "cake", "chair", "sofa", "pottedplant",
"bed", "diningtable", "toilet", "tvmonitor", "laptop", "mouse",
"remote", "keyboard", "cell phone", "microwave", "oven", "toaster",
"sink", "refrigerator", "book", "clock", "vase", "scissors",
"teddy bear", "hair drier", "toothbrush"};
const bool keep_ratio = true;
//vector<int> input_image_; // 输入图片
vector<float> input_image_; // 输入图片
void normalize_(Mat img); // 归一化函数
void nms(vector<BoxInfo>& input_boxes);
Mat resize_image(Mat srcimg, int *newh, int *neww, int *top, int *left);
//Env env = Env(ORT_LOGGING_LEVEL_ERROR, "yolov5-6.1"); // 初始化环境
Env env = Env(ORT_LOGGING_LEVEL_ERROR, "infer"); // 初始化环境
Session *ort_session = nullptr; // 初始化Session指针选项
SessionOptions sessionOptions = SessionOptions(); //初始化Session对象
//SessionOptions sessionOptions;
vector<char*> input_names; // 定义一个字符指针vector
vector<char*> output_names; // 定义一个字符指针vector
vector<vector<int64_t>> input_node_dims; // >=1 outputs ，二维vector
vector<vector<int64_t>> output_node_dims; // >=1 outputs ,int64_t C/C++标准
};
YOLOv5::YOLOv5(Configuration config)
{
this->confThreshold = config.confThreshold;
this->nmsThreshold = config.nmsThreshold;
this->objThreshold = config.objThreshold;
this->num_classes = sizeof(this->classes)/sizeof(this->classes[0]); // 类别数量
this->inpHeight = 640;
this->inpWidth = 640;
string model_path = config.modelpath;
//std::wstring widestr = std::wstring(model_path.begin(), model_path.end()); //用于UTF-16编码的字符
//gpu, https://blog.csdn.net/weixin_44684139/article/details/123504222
//CUDA加速开启
//OrtSessionOptionsAppendExecutionProvider_CUDA(sessionOptions, 0);
//OrtSessionOptionsAppendExecutionProvider_Tensorrt(sessionOptions, 0);
//Ort::ThrowOnError(OrtSessionOptionsAppendExecutionProvider_Tensorrt(sessionOptions, 0));
//Ort::ThrowOnError(OrtSessionOptionsAppendExecutionProvider_CUDA(sessionOptions, 0));
OrtTensorRTProviderOptions trt_options{};
trt_options.trt_max_workspace_size = 2147483648;
trt_options.trt_max_partition_iterations = 10;
trt_options.trt_min_subgraph_size = 5;
trt_options.trt_fp16_enable = 0;
trt_options.trt_int8_enable = 1;
trt_options.trt_int8_use_native_calibration_table = 0;
trt_options.trt_engine_cache_enable = 1;
//trt_options.trt_engine_cache_path = "cache"
trt_options.trt_dump_subgraphs = 1;
sessionOptions.AppendExecutionProvider_TensorRT(trt_options);
// const auto& api = Ort::GetApi();
// OrtTensorRTProviderOptionsV2* tensorrt_options = nullptr;
// api.CreateTensorRTProviderOptions(&tensorrt_options);
// std::unique_ptr<OrtTensorRTProviderOptionsV2, decltype(api.ReleaseTensorRTProviderOptions)> rel_trt_options(
// tensorrt_options, api.ReleaseTensorRTProviderOptions);
// std::vector<const char*> keys{"device_id", "trt_fp16_enable", "trt_int8_enable", "trt_engine_cache_enable","trt_engine_cache_path"};
// std::vector<const char*> values{"0", "1", "0", "1","trt_engine_cache_path"};
// api.UpdateTensorRTProviderOptions(rel_trt_options.get(), keys.data(), values.data(), keys.size());
// api.SessionOptionsAppendExecutionProvider_TensorRT_V2(static_cast<OrtSessionOptions*>(sessionOptions),
// rel_trt_options.get());
std::vector<std::string> availableProviders = Ort::GetAvailableProviders();
std::cout << "availableProviders = " << availableProviders[0] << std::endl;
//sessionOptions.SetGraphOptimizationLevel(ORT_ENABLE_BASIC); //设置图优化类型
sessionOptions.SetGraphOptimizationLevel(ORT_ENABLE_ALL); //设置图优化类型
//ort_session = new Session(env, widestr.c_str(), sessionOptions); // 创建会话，把模型加载到内存中
//ort_session = new Session(env, (const ORTCHAR_T*)model_path.c_str(), sessionOptions); // 创建会话，把模型加载到内存中
ort_session = new Session(env, (const char*)model_path.c_str(), sessionOptions);
size_t numInputNodes = ort_session->GetInputCount(); //输入输出节点数量
size_t numOutputNodes = ort_session->GetOutputCount();
AllocatorWithDefaultOptions allocator; // 配置输入输出节点内存
for (int i = 0; i < numInputNodes; i++)
{
input_names.push_back(ort_session->GetInputName(i, allocator)); // 内存
//Ort::AllocatedStringPtr input_name_Ptr = ort_session->GetInputNameAllocated(i, allocator);
//input_names.push_back(input_name_Ptr.get());
Ort::TypeInfo input_type_info = ort_session->GetInputTypeInfo(i); // 类型
auto input_tensor_info = input_type_info.GetTensorTypeAndShapeInfo(); //
auto input_dims = input_tensor_info.GetShape(); // 输入shape
input_node_dims.push_back(input_dims); // 保存
}
for (int i = 0; i < numOutputNodes; i++)
{
output_names.push_back(ort_session->GetOutputName(i, allocator));
//Ort::AllocatedStringPtr output_name_Ptr = ort_session->GetOutputNameAllocated(i, allocator);
//output_names.push_back(output_name_Ptr.get());
Ort::TypeInfo output_type_info = ort_session->GetOutputTypeInfo(i);
auto output_tensor_info = output_type_info.GetTensorTypeAndShapeInfo();
auto output_dims = output_tensor_info.GetShape();
output_node_dims.push_back(output_dims);
}
this->inpHeight = input_node_dims[0][2];
this->inpWidth = input_node_dims[0][3];
this->nout = output_node_dims[0][2]; // 5+classes
this->num_proposal = output_node_dims[0][1]; // pre_box
}
Mat YOLOv5::resize_image(Mat srcimg, int *newh, int *neww, int *top, int *left)
{
int srch = srcimg.rows, srcw = srcimg.cols;
*newh = this->inpHeight;
*neww = this->inpWidth;
Mat dstimg;
if (this->keep_ratio && srch != srcw) {
float hw_scale = (float)srch / srcw;
if (hw_scale > 1) {
*newh = this->inpHeight;
*neww = int(this->inpWidth / hw_scale);
resize(srcimg, dstimg, Size(*neww, *newh), INTER_AREA);
*left = int((this->inpWidth - *neww) * 0.5);
copyMakeBorder(dstimg, dstimg, 0, 0, *left, this->inpWidth - *neww - *left, BORDER_CONSTANT, 114);
}
else {
*newh = (int)this->inpHeight * hw_scale;
*neww = this->inpWidth;
resize(srcimg, dstimg, Size(*neww, *newh), INTER_AREA);
*top = (int)(this->inpHeight - *newh) * 0.5;
copyMakeBorder(dstimg, dstimg, *top, this->inpHeight - *newh - *top, 0, 0, BORDER_CONSTANT, 114);
}
}
else {
resize(srcimg, dstimg, Size(*neww, *newh), INTER_AREA);
}
return dstimg;
}
void YOLOv5::normalize_(Mat img)
{
// img.convertTo(img, CV_32F);
int row = img.rows;
int col = img.cols;
this->input_image_.resize(row * col * img.channels()); // vector大小
for (int c = 0; c < 3; c++) // bgr
{
for (int i = 0; i < row; i++) // 行
{
for (int j = 0; j < col; j++) // 列
{
float pix = img.ptr<uchar>(i)[j * 3 + 2 - c]; // Mat里的ptr函数访问任意一行像素的首地址,2-c:表示rgb
this->input_image_[c * row * col + i * col + j] = pix / 255.0;
}
}
}
}
void YOLOv5::nms(vector<BoxInfo>& input_boxes)
{
sort(input_boxes.begin(), input_boxes.end(), [](BoxInfo a, BoxInfo b) { return a.score > b.score; }); // 降序排列
vector<float> vArea(input_boxes.size());
for (int i = 0; i < input_boxes.size(); ++i)
{
vArea[i] = (input_boxes[i].x2 - input_boxes[i].x1 + 1)
* (input_boxes[i].y2 - input_boxes[i].y1 + 1);
}
// 全初始化为false，用来作为记录是否保留相应索引下pre_box的标志vector
vector<bool> isSuppressed(input_boxes.size(), false);
for (int i = 0; i < input_boxes.size(); ++i)
{
if (isSuppressed[i]) { continue; }
for (int j = i + 1; j < input_boxes.size(); ++j)
{
if (isSuppressed[j]) { continue; }
float xx1 = max(input_boxes[i].x1, input_boxes[j].x1);
float yy1 = max(input_boxes[i].y1, input_boxes[j].y1);
float xx2 = min(input_boxes[i].x2, input_boxes[j].x2);
float yy2 = min(input_boxes[i].y2, input_boxes[j].y2);
float w = max(0.0f, xx2 - xx1 + 1);
float h = max(0.0f, yy2 - yy1 + 1);
float inter = w * h; // 交集
if(input_boxes[i].label == input_boxes[j].label)
{
float ovr = inter / (vArea[i] + vArea[j] - inter); // 计算iou
if (ovr >= this->nmsThreshold)
{
isSuppressed[j] = true;
}
}
}
}
// return post_nms;
int idx_t = 0;
// remove_if()函数 remove_if(beg, end, op) //移除区间[beg,end)中每一个“令判断式:op(elem)获得true”的元素
input_boxes.erase(remove_if(input_boxes.begin(), input_boxes.end(), [&idx_t, &isSuppressed](const BoxInfo& f) { return isSuppressed[idx_t++]; }), input_boxes.end());
// 另一种写法
// sort(input_boxes.begin(), input_boxes.end(), [](BoxInfo a, BoxInfo b) { return a.score > b.score; }); // 降序排列
// vector<bool> remove_flags(input_boxes.size(),false);
// auto iou = [](const BoxInfo& box1,const BoxInfo& box2)
// {
// float xx1 = max(box1.x1, box2.x1);
// float yy1 = max(box1.y1, box2.y1);
// float xx2 = min(box1.x2, box2.x2);
// float yy2 = min(box1.y2, box2.y2);
// // 交集
// float w = max(0.0f, xx2 - xx1 + 1);
// float h = max(0.0f, yy2 - yy1 + 1);
// float inter_area = w * h;
// // 并集
// float union_area = max(0.0f,box1.x2-box1.x1) * max(0.0f,box1.y2-box1.y1)
// + max(0.0f,box2.x2-box2.x1) * max(0.0f,box2.y2-box2.y1) - inter_area;
// return inter_area / union_area;
// };
// for (int i = 0; i < input_boxes.size(); ++i)
// {
// if(remove_flags[i]) continue;
// for (int j = i + 1; j < input_boxes.size(); ++j)
// {
// if(remove_flags[j]) continue;
// if(input_boxes[i].label == input_boxes[j].label && iou(input_boxes[i],input_boxes[j])>=this->nmsThreshold)
// {
// remove_flags[j] = true;
// }
// }
// }
// int idx_t = 0;
// // remove_if()函数 remove_if(beg, end, op) //移除区间[beg,end)中每一个“令判断式:op(elem)获得true”的元素
// input_boxes.erase(remove_if(input_boxes.begin(), input_boxes.end(), [&idx_t, &remove_flags](const BoxInfo& f) { return remove_flags[idx_t++]; }), input_boxes.end());
}
void YOLOv5::detect(Mat& frame)
{
int newh = 0, neww = 0, padh = 0, padw = 0;
Mat dstimg = this->resize_image(frame, &newh, &neww, &padh, &padw);
this->normalize_(dstimg);
// 定义一个输入矩阵，int64_t是下面作为输入参数时的类型
array<int64_t, 4> input_shape_{ 1, 3, this->inpHeight, this->inpWidth };
//创建输入tensor
auto allocator_info = MemoryInfo::CreateCpu(OrtDeviceAllocator, OrtMemTypeCPU);
Value input_tensor_ = Value::CreateTensor<float>(allocator_info, input_image_.data(), input_image_.size(), input_shape_.data(), input_shape_.size());
//Value input_tensor_ = Value::CreateTensor<int>(allocator_info, input_image_.data(), input_image_.size(), input_shape_.data(), input_shape_.size());
// 开始推理
vector<Value> ort_outputs = ort_session->Run(RunOptions{ nullptr }, &input_names[0], &input_tensor_, 1, output_names.data(), output_names.size()); // 开始推理
/generate proposals
vector<BoxInfo> generate_boxes; // BoxInfo自定义的结构体
float ratioh = (float)frame.rows / newh, ratiow = (float)frame.cols / neww;
float* pdata = ort_outputs[0].GetTensorMutableData<float>(); // GetTensorMutableData
for(int i = 0; i < num_proposal; ++i) // 遍历所有的num_pre_boxes
{
int index = i * nout; // prob[b*num_pred_boxes*(classes+5)]
float obj_conf = pdata[index + 4]; // 置信度分数
if (obj_conf > this->objThreshold) // 大于阈值
{
int class_idx = 0;
float max_class_socre = 0;
for (int k = 0; k < this->num_classes; ++k)
{
if (pdata[k + index + 5] > max_class_socre)
{
max_class_socre = pdata[k + index + 5];
class_idx = k;
}
}
max_class_socre *= obj_conf; // 最大的类别分数*置信度
if (max_class_socre > this->confThreshold) // 再次筛选
{
//const int class_idx = classIdPoint.x;
float cx = pdata[index]; //x
float cy = pdata[index+1]; //y
float w = pdata[index+2]; //w
float h = pdata[index+3]; //h
float xmin = (cx - padw - 0.5 * w)*ratiow;
float ymin = (cy - padh - 0.5 * h)*ratioh;
float xmax = (cx - padw + 0.5 * w)*ratiow;
float ymax = (cy - padh + 0.5 * h)*ratioh;
generate_boxes.push_back(BoxInfo{ xmin, ymin, xmax, ymax, max_class_socre, class_idx });
}
}
}
// Perform non maximum suppression to eliminate redundant overlapping boxes with
// lower confidences
nms(generate_boxes);
for (size_t i = 0; i < generate_boxes.size(); ++i)
{
int xmin = int(generate_boxes[i].x1);
int ymin = int(generate_boxes[i].y1);
rectangle(frame, Point(xmin, ymin), Point(int(generate_boxes[i].x2), int(generate_boxes[i].y2)), Scalar(0, 0, 255), 2);
string label = format("%.2f", generate_boxes[i].score);
label = this->classes[generate_boxes[i].label] + ":" + label;
putText(frame, label, Point(xmin, ymin - 5), FONT_HERSHEY_SIMPLEX, 0.75, Scalar(0, 255, 0), 1);
}
}
int main(int argc,char *argv[])
{
clock_t startTime,endTime; //计算时间
Configuration yolo_nets = { 0.3, 0.5, 0.3,"/home/jingjia/code/ken/yolov5-onnxruntime-master/models1/yolov5s_int8q_jingtai_1.onnx" };
//Configuration yolo_nets = { 0.3, 0.5, 0.3,"/home/jingjia/code/ken/yolov5-onnxruntime-master/models/yolov5s.onnx" };
//Configuration yolo_nets = { 0.3, 0.5, 0.3,"/home/jingjia/code/ken/yolov5_test/TensorrtExecutionProvider_TRT_Subgraph.onnx" };
YOLOv5 yolo_model(yolo_nets);
string imgpath = "/home/jingjia/code/ken/yolov5-onnxruntime-master/images/bus.jpg";
Mat srcimg = imread(imgpath);
double timeStart = (double)getTickCount();
startTime = clock();//计时开始
yolo_model.detect(srcimg);
endTime = clock();//计时结束
double nTime = ((double)getTickCount() - timeStart) / getTickFrequency();
cout << "clock_running time is:" <<(double)(endTime - startTime) / CLOCKS_PER_SEC << "s" << endl;
cout << "The run time is:" << (double)clock() /CLOCKS_PER_SEC<< "s" << endl;
cout << "getTickCount_running time :" << nTime << "sec\n" << endl;
// static const string kWinName = "Deep learning object detection in ONNXRuntime";
// namedWindow(kWinName, WINDOW_NORMAL);
// imshow(kWinName, srcimg);
imwrite("restult_ort.jpg",srcimg);
// waitKey(0);
// destroyAllWindows();
return 0;
}?