Windows C++ VS2022 OpenVINO 实例分割 Demo

  • 开发
  • 60

目录

效果

模型信息

项目

代码

下载

其他

Windows C++ VS2022 OpenVINO 实例分割 Demo

效果

模型信息

Model Properties
-------------------------
date：2023-09-07T17:11:46.798385
description：Ultralytics YOLOv8n-seg model trained on coco.yaml
author：Ultralytics
task：segment
license：AGPL-3.0 https://ultralytics.com/license
version：8.0.172
stride：32
batch：1
imgsz：[640, 640]
names：{0: 'person', 1: 'bicycle', 2: 'car', 3: 'motorcycle', 4: 'airplane', 5: 'bus', 6: 'train', 7: 'truck', 8: 'boat', 9: 'traffic light', 10: 'fire hydrant', 11: 'stop sign', 12: 'parking meter', 13: 'bench', 14: 'bird', 15: 'cat', 16: 'dog', 17: 'horse', 18: 'sheep', 19: 'cow', 20: 'elephant', 21: 'bear', 22: 'zebra', 23: 'giraffe', 24: 'backpack', 25: 'umbrella', 26: 'handbag', 27: 'tie', 28: 'suitcase', 29: 'frisbee', 30: 'skis', 31: 'snowboard', 32: 'sports ball', 33: 'kite', 34: 'baseball bat', 35: 'baseball glove', 36: 'skateboard', 37: 'surfboard', 38: 'tennis racket', 39: 'bottle', 40: 'wine glass', 41: 'cup', 42: 'fork', 43: 'knife', 44: 'spoon', 45: 'bowl', 46: 'banana', 47: 'apple', 48: 'sandwich', 49: 'orange', 50: 'broccoli', 51: 'carrot', 52: 'hot dog', 53: 'pizza', 54: 'donut', 55: 'cake', 56: 'chair', 57: 'couch', 58: 'potted plant', 59: 'bed', 60: 'dining table', 61: 'toilet', 62: 'tv', 63: 'laptop', 64: 'mouse', 65: 'remote', 66: 'keyboard', 67: 'cell phone', 68: 'microwave', 69: 'oven', 70: 'toaster', 71: 'sink', 72: 'refrigerator', 73: 'book', 74: 'clock', 75: 'vase', 76: 'scissors', 77: 'teddy bear', 78: 'hair drier', 79: 'toothbrush'}
---------------------------------------------------------------

Inputs
-------------------------
name：images
tensor：Float[1, 3, 640, 640]
---------------------------------------------------------------

Outputs
-------------------------
name：output0
tensor：Float[1, 116, 8400]
name：output1
tensor：Float[1, 32, 160, 160]
---------------------------------------------------------------

项目

代码

#include <iostream>
#include <string>
#include <vector>
#include <algorithm>
 
#include <openvino/openvino.hpp> //openvino header file
#include <opencv2/opencv.hpp>    //opencv header file
 
#include  <direct.h>  
#include  <stdio.h> 
 
using namespace cv;
using namespace dnn;
 
std::vector<Scalar> colors = { Scalar(255, 0, 0), Scalar(255, 0, 255), Scalar(170, 0, 255), Scalar(255, 0, 85),
                                   Scalar(255, 0, 170), Scalar(85, 255, 0), Scalar(255, 170, 0), Scalar(0, 255, 0),
                                   Scalar(255, 255, 0), Scalar(0, 255, 85), Scalar(170, 255, 0), Scalar(0, 85, 255),
                                   Scalar(0, 255, 170), Scalar(0, 0, 255), Scalar(0, 255, 255), Scalar(85, 0, 255) };
 
const std::vector<std::string> class_names = {
    "person", "bicycle", "car", "motorcycle", "airplane", "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", "couch",
    "potted plant", "bed", "dining table", "toilet", "tv", "laptop", "mouse", "remote", "keyboard", "cell phone",
    "microwave", "oven", "toaster", "sink", "refrigerator", "book", "clock", "vase", "scissors", "teddy bear",
    "hair drier", "toothbrush" };
 
// Keep the ratio before resize
Mat letterbox(const cv::Mat& source)
{
    int col = source.cols;
    int row = source.rows;
    int _max = MAX(col, row);
    Mat result = Mat::zeros(_max, _max, CV_8UC3);
    source.copyTo(result(Rect(0, 0, col, row)));
    return result;
}
 
float sigmoid_function(float a) {
    float b = 1. / (1. + exp(-a));
    return b;
}
 
int main(int argc, char* argv[])
{
 
    char   buffer[100];
    _getcwd(buffer, 100);
    std::cout << "当前路径:" << buffer << std::endl;
 
    // -------- Step 1. Initialize OpenVINO Runtime Core --------
    ov::Core core;
 
    // -------- Step 2. Compile the Model --------
    String model_path = String(buffer) + "\\yolov8n-seg.xml";
    auto compiled_model = core.compile_model(model_path, "CPU");
 
    // -------- Step 3. Create an Inference Request --------
    ov::InferRequest infer_request = compiled_model.create_infer_request();
 
    // -------- Step 4.Read a picture file and do the preprocess --------
    Mat img = cv::imread("bus.jpg");
    // Preprocess the image
    Mat letterbox_img = letterbox(img);
    float scale = letterbox_img.size[0] / 640.0;
    Mat blob = blobFromImage(letterbox_img, 1.0 / 255.0, Size(640, 640), Scalar(), true);
 
    // -------- Step 5. Feed the blob into the input node of the Model -------
    // Get input port for model with one input
    auto input_port = compiled_model.input();
    // Create tensor from external memory
    ov::Tensor input_tensor(input_port.get_element_type(), input_port.get_shape(), blob.ptr(0));
    // Set input tensor for model with one input
    infer_request.set_input_tensor(input_tensor);
 
    // -------- Step 6. Start inference --------
    infer_request.infer();
 
    // -------- Step 7. Get the inference result --------
    auto output0 = infer_request.get_output_tensor(0); //output0
    auto output1 = infer_request.get_output_tensor(1); //otuput1
    auto output0_shape = output0.get_shape();
    auto output1_shape = output1.get_shape();
    std::cout << "The shape of output0:" << output0_shape << std::endl;
    std::cout << "The shape of output1:" << output1_shape << std::endl;
 
    // -------- Step 8. Postprocess the result --------
    Mat output_buffer(output0_shape[1], output0_shape[2], CV_32F, output0.data<float>());
    Mat proto(32, 25600, CV_32F, output1.data<float>()); //[32,25600]
    transpose(output_buffer, output_buffer); //[8400,116]
    float score_threshold = 0.25;
    float nms_threshold = 0.5;
    std::vector<int> class_ids;
    std::vector<float> class_scores;
    std::vector<Rect> boxes;
    std::vector<Mat> mask_confs;
    // Figure out the bbox, class_id and class_score
    for (int i = 0; i < output_buffer.rows; i++) {
        Mat classes_scores = output_buffer.row(i).colRange(4, 84);
        Point class_id;
        double maxClassScore;
        minMaxLoc(classes_scores, 0, &maxClassScore, 0, &class_id);
 
        if (maxClassScore > score_threshold) {
            class_scores.push_back(maxClassScore);
            class_ids.push_back(class_id.x);
            float cx = output_buffer.at<float>(i, 0);
            float cy = output_buffer.at<float>(i, 1);
            float w = output_buffer.at<float>(i, 2);
            float h = output_buffer.at<float>(i, 3);
 
            int left = int((cx - 0.5 * w) * scale);
            int top = int((cy - 0.5 * h) * scale);
            int width = int(w * scale);
            int height = int(h * scale);
 
            cv::Mat mask_conf = output_buffer.row(i).colRange(84, 116);
            mask_confs.push_back(mask_conf);
            boxes.push_back(Rect(left, top, width, height));
        }
    }
    //NMS
    std::vector<int> indices;
    NMSBoxes(boxes, class_scores, score_threshold, nms_threshold, indices);
 
    // -------- Visualize the detection results -----------
    cv::Mat rgb_mask = cv::Mat::zeros(img.size(), CV_8UC3);
    cv::Mat masked_img;
    cv::RNG rng;
 
    for (size_t i = 0; i < indices.size(); i++) {
        // Visualize the objects
        int index = indices[i];
        int class_id = class_ids[index];
        rectangle(img, boxes[index], colors[class_id % 16], 2, 8);
        std::string label = class_names[class_id] + ":" + std::to_string(class_scores[index]).substr(0, 4);
        Size textSize = cv::getTextSize(label, FONT_HERSHEY_SIMPLEX, 0.5, 1, 0);
        Rect textBox(boxes[index].tl().x, boxes[index].tl().y - 15, textSize.width, textSize.height + 5);
        cv::rectangle(img, textBox, colors[class_id % 16], FILLED);
        putText(img, label, Point(boxes[index].tl().x, boxes[index].tl().y - 5), FONT_HERSHEY_SIMPLEX, 0.5, Scalar(255, 255, 255));
 
        // Visualize the Masks
        Mat m = mask_confs[index] * proto;
        for (int col = 0; col < m.cols; col++) {
            m.at<float>(0, col) = sigmoid_function(m.at<float>(0, col));
        }
 
        cv::Mat m1 = m.reshape(1, 160); // 1x25600 -> 160x160
 
        int x1 = std::max(0, boxes[index].x);
        int y1 = std::max(0, boxes[index].y);
        int x2 = std::max(0, boxes[index].br().x);
        int y2 = std::max(0, boxes[index].br().y);
 
        int mx1 = int(x1 / scale * 0.25);
        int my1 = int(y1 / scale * 0.25);
        int mx2 = int(x2 / scale * 0.25);
        int my2 = int(y2 / scale * 0.25);
 
        cv::Mat mask_roi = m1(cv::Range(my1, my2), cv::Range(mx1, mx2));
        cv::Mat rm, det_mask;
        cv::resize(mask_roi, rm, cv::Size(x2 - x1, y2 - y1));
 
        for (int r = 0; r < rm.rows; r++) {
            for (int c = 0; c < rm.cols; c++) {
                float pv = rm.at<float>(r, c);
                if (pv > 0.5) {
                    rm.at<float>(r, c) = 1.0;
                }
                else {
                    rm.at<float>(r, c) = 0.0;
                }
            }
        }
        rm = rm * rng.uniform(0, 255);
        rm.convertTo(det_mask, CV_8UC1);
        if ((y1 + det_mask.rows) >= img.rows) {
            y2 = img.rows - 1;
        }
        if ((x1 + det_mask.cols) >= img.cols) {
            x2 = img.cols - 1;
        }
 
        cv::Mat mask = cv::Mat::zeros(cv::Size(img.cols, img.rows), CV_8UC1);
       
        det_mask= det_mask( cv::Range(0, y2 - y1), cv::Range(0, x2 - x1));
 
        Rect roi(x1, y1, x2 - x1, y2 - y1);
        det_mask.copyTo(Mat(mask, roi));
        
        add(rgb_mask, cv::Scalar(rng.uniform(0, 255), rng.uniform(0, 255), rng.uniform(0, 255)), rgb_mask, mask);
        
        addWeighted(img, 0.5, rgb_mask, 0.5, 0, masked_img);
    }
 
    namedWindow("YOLOv8-Seg OpenVINO Inference C++ Demo", WINDOW_AUTOSIZE);
    imshow("YOLOv8-Seg OpenVINO Inference C++ Demo", masked_img);
    waitKey(0);
    destroyAllWindows();
    return 0;
}

#include <iostream>
#include <string>
#include <vector>
#include <algorithm>
 
#include <openvino/openvino.hpp> //openvino header file
#include <opencv2/opencv.hpp>    //opencv header file
 
#include  <direct.h>  
#include  <stdio.h> 
 
using namespace cv;
using namespace dnn;
 
std::vector<Scalar> colors = { Scalar(255, 0, 0), Scalar(255, 0, 255), Scalar(170, 0, 255), Scalar(255, 0, 85),
                                   Scalar(255, 0, 170), Scalar(85, 255, 0), Scalar(255, 170, 0), Scalar(0, 255, 0),
                                   Scalar(255, 255, 0), Scalar(0, 255, 85), Scalar(170, 255, 0), Scalar(0, 85, 255),
                                   Scalar(0, 255, 170), Scalar(0, 0, 255), Scalar(0, 255, 255), Scalar(85, 0, 255) };
 
const std::vector<std::string> class_names = {
    "person", "bicycle", "car", "motorcycle", "airplane", "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", "couch",
    "potted plant", "bed", "dining table", "toilet", "tv", "laptop", "mouse", "remote", "keyboard", "cell phone",
    "microwave", "oven", "toaster", "sink", "refrigerator", "book", "clock", "vase", "scissors", "teddy bear",
    "hair drier", "toothbrush" };
 
// Keep the ratio before resize
Mat letterbox(const cv::Mat& source)
{
    int col = source.cols;
    int row = source.rows;
    int _max = MAX(col, row);
    Mat result = Mat::zeros(_max, _max, CV_8UC3);
    source.copyTo(result(Rect(0, 0, col, row)));
    return result;
}
 
float sigmoid_function(float a) {
    float b = 1. / (1. + exp(-a));
    return b;
}
 
int main(int argc, char* argv[])
{
 
    char   buffer[100];
    _getcwd(buffer, 100);
    std::cout << "当前路径:" << buffer << std::endl;
 
    // -------- Step 1. Initialize OpenVINO Runtime Core --------
    ov::Core core;
 
    // -------- Step 2. Compile the Model --------
    String model_path = String(buffer) + "\\yolov8n-seg.xml";
    auto compiled_model = core.compile_model(model_path, "CPU");
 
    // -------- Step 3. Create an Inference Request --------
    ov::InferRequest infer_request = compiled_model.create_infer_request();
 
    // -------- Step 4.Read a picture file and do the preprocess --------
    Mat img = cv::imread("bus.jpg");
    // Preprocess the image
    Mat letterbox_img = letterbox(img);
    float scale = letterbox_img.size[0] / 640.0;
    Mat blob = blobFromImage(letterbox_img, 1.0 / 255.0, Size(640, 640), Scalar(), true);
 
    // -------- Step 5. Feed the blob into the input node of the Model -------
    // Get input port for model with one input
    auto input_port = compiled_model.input();
    // Create tensor from external memory
    ov::Tensor input_tensor(input_port.get_element_type(), input_port.get_shape(), blob.ptr(0));
    // Set input tensor for model with one input
    infer_request.set_input_tensor(input_tensor);
 
    // -------- Step 6. Start inference --------
    infer_request.infer();
 
    // -------- Step 7. Get the inference result --------
    auto output0 = infer_request.get_output_tensor(0); //output0
    auto output1 = infer_request.get_output_tensor(1); //otuput1
    auto output0_shape = output0.get_shape();
    auto output1_shape = output1.get_shape();
    std::cout << "The shape of output0:" << output0_shape << std::endl;
    std::cout << "The shape of output1:" << output1_shape << std::endl;
 
    // -------- Step 8. Postprocess the result --------
    Mat output_buffer(output0_shape[1], output0_shape[2], CV_32F, output0.data<float>());
    Mat proto(32, 25600, CV_32F, output1.data<float>()); //[32,25600]
    transpose(output_buffer, output_buffer); //[8400,116]
    float score_threshold = 0.25;
    float nms_threshold = 0.5;
    std::vector<int> class_ids;
    std::vector<float> class_scores;
    std::vector<Rect> boxes;
    std::vector<Mat> mask_confs;
    // Figure out the bbox, class_id and class_score
    for (int i = 0; i < output_buffer.rows; i++) {
        Mat classes_scores = output_buffer.row(i).colRange(4, 84);
        Point class_id;
        double maxClassScore;
        minMaxLoc(classes_scores, 0, &maxClassScore, 0, &class_id);
 
        if (maxClassScore > score_threshold) {
            class_scores.push_back(maxClassScore);
            class_ids.push_back(class_id.x);
            float cx = output_buffer.at<float>(i, 0);
            float cy = output_buffer.at<float>(i, 1);
            float w = output_buffer.at<float>(i, 2);
            float h = output_buffer.at<float>(i, 3);
 
            int left = int((cx - 0.5 * w) * scale);
            int top = int((cy - 0.5 * h) * scale);
            int width = int(w * scale);
            int height = int(h * scale);
 
            cv::Mat mask_conf = output_buffer.row(i).colRange(84, 116);
            mask_confs.push_back(mask_conf);
            boxes.push_back(Rect(left, top, width, height));
        }
    }
    //NMS
    std::vector<int> indices;
    NMSBoxes(boxes, class_scores, score_threshold, nms_threshold, indices);
 
    // -------- Visualize the detection results -----------
    cv::Mat rgb_mask = cv::Mat::zeros(img.size(), CV_8UC3);
    cv::Mat masked_img;
    cv::RNG rng;
 
    for (size_t i = 0; i < indices.size(); i++) {
        // Visualize the objects
        int index = indices[i];
        int class_id = class_ids[index];
        rectangle(img, boxes[index], colors[class_id % 16], 2, 8);
        std::string label = class_names[class_id] + ":" + std::to_string(class_scores[index]).substr(0, 4);
        Size textSize = cv::getTextSize(label, FONT_HERSHEY_SIMPLEX, 0.5, 1, 0);
        Rect textBox(boxes[index].tl().x, boxes[index].tl().y - 15, textSize.width, textSize.height + 5);
        cv::rectangle(img, textBox, colors[class_id % 16], FILLED);
        putText(img, label, Point(boxes[index].tl().x, boxes[index].tl().y - 5), FONT_HERSHEY_SIMPLEX, 0.5, Scalar(255, 255, 255));
 
        // Visualize the Masks
        Mat m = mask_confs[index] * proto;
        for (int col = 0; col < m.cols; col++) {
            m.at<float>(0, col) = sigmoid_function(m.at<float>(0, col));
        }
 
        cv::Mat m1 = m.reshape(1, 160); // 1x25600 -> 160x160
 
        int x1 = std::max(0, boxes[index].x);
        int y1 = std::max(0, boxes[index].y);
        int x2 = std::max(0, boxes[index].br().x);
        int y2 = std::max(0, boxes[index].br().y);
 
        int mx1 = int(x1 / scale * 0.25);
        int my1 = int(y1 / scale * 0.25);
        int mx2 = int(x2 / scale * 0.25);
        int my2 = int(y2 / scale * 0.25);
 
        cv::Mat mask_roi = m1(cv::Range(my1, my2), cv::Range(mx1, mx2));
        cv::Mat rm, det_mask;
        cv::resize(mask_roi, rm, cv::Size(x2 - x1, y2 - y1));
 
        for (int r = 0; r < rm.rows; r++) {
            for (int c = 0; c < rm.cols; c++) {
                float pv = rm.at<float>(r, c);
                if (pv > 0.5) {
                    rm.at<float>(r, c) = 1.0;
                }
                else {
                    rm.at<float>(r, c) = 0.0;
                }
            }
        }
        rm = rm * rng.uniform(0, 255);
        rm.convertTo(det_mask, CV_8UC1);
        if ((y1 + det_mask.rows) >= img.rows) {
            y2 = img.rows - 1;
        }
        if ((x1 + det_mask.cols) >= img.cols) {
            x2 = img.cols - 1;
        }
 
        cv::Mat mask = cv::Mat::zeros(cv::Size(img.cols, img.rows), CV_8UC1);
       
        det_mask= det_mask( cv::Range(0, y2 - y1), cv::Range(0, x2 - x1));
 
        Rect roi(x1, y1, x2 - x1, y2 - y1);
        det_mask.copyTo(Mat(mask, roi));
        
        add(rgb_mask, cv::Scalar(rng.uniform(0, 255), rng.uniform(0, 255), rng.uniform(0, 255)), rgb_mask, mask);
        
        addWeighted(img, 0.5, rgb_mask, 0.5, 0, masked_img);
    }
 
    namedWindow("YOLOv8-Seg OpenVINO Inference C++ Demo", WINDOW_AUTOSIZE);
    imshow("YOLOv8-Seg OpenVINO Inference C++ Demo", masked_img);
    waitKey(0);
    destroyAllWindows();
    return 0;
}

下载

 源码下载

其他

环境配置参考：https://lw112190.blog.csdn.net/article/details/132827809

阅读全部

相关推荐

  6. CSS3DObject 分子公式demo 实现

    2023-12-18 14:00:02       38 阅读

最近更新

  3. docker php8.1+nginx base 镜像 dockerfile 配置

    2023-12-18 14:00:02       98 阅读

  4. Could not load dynamic library ‘cudart64_100.dll‘

    2023-12-18 14:00:02       106 阅读

  9. 在Django里面运行非项目文件

    2023-12-18 14:00:02       87 阅读

  19. Python语言-面向对象

    2023-12-18 14:00:02       96 阅读

  20. 如何分清楚常见的 Git 分支管理策略Git Flow、GitHub Flow 和 GitLab Flow

    2023-12-18 14:00:02       91 阅读

热门阅读

  3. Rust语言中match详解

    2023-12-18 14:00:02       63 阅读

  4. 移动零算法(leetcode第283题)

    2023-12-18 14:00:02       55 阅读

  17. 【INTEL（ALTERA）】quartus报错UVM_FATAL [cxl_tb_top_initialize] Gen5 链接失败。超时怎么办

    2023-12-18 14:00:02       61 阅读

  20. springboot 2.7.0 集成 swagger

    2023-12-18 14:00:02       55 阅读

  21. Python3 数字(Number) ----20231215

    2023-12-18 14:00:02       61 阅读

  25. Powershell summaries with types of scales of summaries

    2023-12-18 14:00:02       63 阅读

  26. Python multiprocessing包的基本使用

    2023-12-18 14:00:02       59 阅读

  29. 设计模式_行为型模式_观察者模式

    2023-12-18 14:00:02       64 阅读