C# OpenVINO 直接读取百度Paddle模型实现物体检测( yolov3_darknet)

效果

项目

代码

using OpenCvSharp;
using Sdcb.OpenVINO;
using Sdcb.OpenVINO.Natives;
using System;
using System.Collections.Generic;
using System.Diagnostics;
using System.Drawing;
using System.IO;
using System.Runtime.InteropServices;
using System.Text;
using System.Windows.Forms;
using System.Xml.Linq;
using System.Xml.XPath;

namespace OpenVINO_Det_物体检测
{
public partial class Form1 : Form
{
public Form1()
{
InitializeComponent();
}

string fileFilter = "*.*|*.bmp;*.jpg;*.jpeg;*.tiff;*.tiff;*.png";
string image_path = "";
string startupPath;
string model_path;
Mat src;
string[] dicts;

StringBuilder sb = new StringBuilder();

float confidence = 0.75f;

private void button1_Click(object sender, EventArgs e)
{
OpenFileDialog ofd = new OpenFileDialog();
ofd.Filter = fileFilter;
if (ofd.ShowDialog() != DialogResult.OK) return;
pictureBox1.Image = null;
image_path = ofd.FileName;
pictureBox1.Image = new Bitmap(image_path);
textBox1.Text = "";
src = new Mat(image_path);
pictureBox2.Image = null;
}

unsafe private void button2_Click(object sender, EventArgs e)
{
if (pictureBox1.Image == null)
{
return;
}

pictureBox2.Image = null;
textBox1.Text = "";
sb.Clear();

src = new Mat(image_path);
Mat result_image = src.Clone();

model_path = "yolov3_darknet/model.pdmodel";
Model rawModel = OVCore.Shared.ReadModel(model_path);

int inpHeight = 608;
int inpWidth = 608;

var ad = OVCore.Shared.AvailableDevices;
Console.WriteLine("可用设备");
foreach (var item in ad)
{
Console.WriteLine(item);
}

CompiledModel cm = OVCore.Shared.CompileModel(rawModel, "CPU");
InferRequest ir = cm.CreateInferRequest();

Stopwatch stopwatch = new Stopwatch();

Shape inputShape = new Shape(1, 608, 608);
Size2f sizeRatio = new Size2f(1f * src.Width / inputShape[2], 1f * src.Height / inputShape[1]);
Cv2.CvtColor(src, src, ColorConversionCodes.BGR2RGB);

Point2f scaleRate = new Point2f(1f * inpWidth / src.Width, 1f * inpHeight / src.Height);

Cv2.Resize(src, src, new OpenCvSharp.Size(), scaleRate.X, scaleRate.Y);

Common.Normalize(src);

float[] input_tensor_data = Common.ExtractMat(src);

/*
scale_factor 1,2
image 1,3,608,608
im_shape 1,2
*/
Tensor input_scale_factor = Tensor.FromArray(new float[] { scaleRate.Y, scaleRate.X }, new Shape(1, 2));
Tensor input_image = Tensor.FromArray(input_tensor_data, new Shape(1, 3, 608, 608));
Tensor input_im_shape = Tensor.FromArray(new float[] { 608, 608 }, new Shape(1, 2));

ir.Inputs[0] = input_scale_factor;
ir.Inputs[1] = input_image;
ir.Inputs[2] = input_im_shape;

double preprocessTime = stopwatch.Elapsed.TotalMilliseconds;
stopwatch.Restart();

ir.Run();

double inferTime = stopwatch.Elapsed.TotalMilliseconds;
stopwatch.Restart();

Tensor output_0 = ir.Outputs[0];

int num = (int)output_0.Shape.Dimensions[0];

float[] output_0_array = output_0.GetData<float>().ToArray();

for (int j = 0; j < num; j++)
{
int num12 = (int)Math.Round(output_0_array[j * 6]);
float score = output_0_array[1 + j * 6];

if (score > this.confidence)
{
int num13 = (int)(output_0_array[2 + j * 6]);
int num14 = (int)(output_0_array[3 + j * 6]);
int num15 = (int)(output_0_array[4 + j * 6]);
int num16 = (int)(output_0_array[5 + j * 6]);

string ClassName = dicts[num12];
Rect r = Rect.FromLTRB(num13, num14, num15, num16);
sb.AppendLine($"{ClassName}:{score:P0}");
Cv2.PutText(result_image, $"{ClassName}:{score:P0}", new OpenCvSharp.Point(r.TopLeft.X, r.TopLeft.Y - 10), HersheyFonts.HersheySimplex, 1, Scalar.Red, 2);
Cv2.Rectangle(result_image, r, Scalar.Red, thickness: 2);
}
}

double postprocessTime = stopwatch.Elapsed.TotalMilliseconds;
stopwatch.Stop();
double totalTime = preprocessTime + inferTime + postprocessTime;

pictureBox2.Image = new Bitmap(result_image.ToMemoryStream());

sb.AppendLine($"Preprocess: {preprocessTime:F2}ms");
sb.AppendLine($"Infer: {inferTime:F2}ms");
sb.AppendLine($"Postprocess: {postprocessTime:F2}ms");
sb.AppendLine($"Total: {totalTime:F2}ms");

textBox1.Text = sb.ToString();

}

private void Form1_Load(object sender, EventArgs e)
{
startupPath = Application.StartupPath;

string classer_path = "lable.txt";
List<string> str = new List<string>();
StreamReader sr = new StreamReader(classer_path);
string line;
while ((line = sr.ReadLine()) != null)
{
str.Add(line);
}
dicts = str.ToArray();

image_path = "1.png";
pictureBox1.Image = new Bitmap(image_path);
}
}
}

using OpenCvSharp;
using Sdcb.OpenVINO;
using Sdcb.OpenVINO.Natives;
using System;
using System.Collections.Generic;
using System.Diagnostics;
using System.Drawing;
using System.IO;
using System.Runtime.InteropServices;
using System.Text;
using System.Windows.Forms;
using System.Xml.Linq;
using System.Xml.XPath;

namespace OpenVINO_Det_物体检测
{
    public partial class Form1 : Form
    {
        public Form1()
        {
            InitializeComponent();
        }

        string fileFilter = "*.*|*.bmp;*.jpg;*.jpeg;*.tiff;*.tiff;*.png";
        string image_path = "";
        string startupPath;
        string model_path;
        Mat src;
        string[] dicts;

        StringBuilder sb = new StringBuilder();

        float confidence = 0.75f;

        private void button1_Click(object sender, EventArgs e)
        {
            OpenFileDialog ofd = new OpenFileDialog();
            ofd.Filter = fileFilter;
            if (ofd.ShowDialog() != DialogResult.OK) return;
            pictureBox1.Image = null;
            image_path = ofd.FileName;
            pictureBox1.Image = new Bitmap(image_path);
            textBox1.Text = "";
            src = new Mat(image_path);
            pictureBox2.Image = null;
        }

        unsafe private void button2_Click(object sender, EventArgs e)
        {
            if (pictureBox1.Image == null)
            {
                return;
            }

            pictureBox2.Image = null;
            textBox1.Text = "";
            sb.Clear();

            src = new Mat(image_path);
            Mat result_image = src.Clone();

            model_path = "yolov3_darknet/model.pdmodel";
            Model rawModel = OVCore.Shared.ReadModel(model_path);

            int inpHeight = 608;
            int inpWidth = 608;

            var ad = OVCore.Shared.AvailableDevices;
            Console.WriteLine("可用设备");
            foreach (var item in ad)
            {
                Console.WriteLine(item);
            }

            CompiledModel cm = OVCore.Shared.CompileModel(rawModel, "CPU");
            InferRequest ir = cm.CreateInferRequest();

            Stopwatch stopwatch = new Stopwatch();

            Shape inputShape = new Shape(1, 608, 608);
            Size2f sizeRatio = new Size2f(1f * src.Width / inputShape[2], 1f * src.Height / inputShape[1]);
            Cv2.CvtColor(src, src, ColorConversionCodes.BGR2RGB);

            Point2f scaleRate = new Point2f(1f * inpWidth / src.Width, 1f * inpHeight / src.Height);

            Cv2.Resize(src, src, new OpenCvSharp.Size(), scaleRate.X, scaleRate.Y);

            Common.Normalize(src);

            float[] input_tensor_data = Common.ExtractMat(src);

            /*
             scale_factor   1,2
             image          1,3,608,608
             im_shape       1,2 
             */
            Tensor input_scale_factor = Tensor.FromArray(new float[] { scaleRate.Y, scaleRate.X }, new Shape(1, 2));
            Tensor input_image = Tensor.FromArray(input_tensor_data, new Shape(1, 3, 608, 608));
            Tensor input_im_shape = Tensor.FromArray(new float[] { 608, 608 }, new Shape(1, 2));

            ir.Inputs[0] = input_scale_factor;
            ir.Inputs[1] = input_image;
            ir.Inputs[2] = input_im_shape;

            double preprocessTime = stopwatch.Elapsed.TotalMilliseconds;
            stopwatch.Restart();

            ir.Run();

            double inferTime = stopwatch.Elapsed.TotalMilliseconds;
            stopwatch.Restart();

            Tensor output_0 = ir.Outputs[0];

            int num = (int)output_0.Shape.Dimensions[0];

            float[] output_0_array = output_0.GetData<float>().ToArray();

            for (int j = 0; j < num; j++)
            {
                int num12 = (int)Math.Round(output_0_array[j * 6]);
                float score = output_0_array[1 + j * 6];

                if (score > this.confidence)
                {
                    int num13 = (int)(output_0_array[2 + j * 6]);
                    int num14 = (int)(output_0_array[3 + j * 6]);
                    int num15 = (int)(output_0_array[4 + j * 6]);
                    int num16 = (int)(output_0_array[5 + j * 6]);

                    string ClassName = dicts[num12];
                    Rect r = Rect.FromLTRB(num13, num14, num15, num16);
                    sb.AppendLine($"{ClassName}:{score:P0}");
                    Cv2.PutText(result_image, $"{ClassName}:{score:P0}", new OpenCvSharp.Point(r.TopLeft.X, r.TopLeft.Y - 10), HersheyFonts.HersheySimplex, 1, Scalar.Red, 2);
                    Cv2.Rectangle(result_image, r, Scalar.Red, thickness: 2);
                }
            }

            double postprocessTime = stopwatch.Elapsed.TotalMilliseconds;
            stopwatch.Stop();
            double totalTime = preprocessTime + inferTime + postprocessTime;

            pictureBox2.Image = new Bitmap(result_image.ToMemoryStream());

            sb.AppendLine($"Preprocess: {preprocessTime:F2}ms");
            sb.AppendLine($"Infer: {inferTime:F2}ms");
            sb.AppendLine($"Postprocess: {postprocessTime:F2}ms");
            sb.AppendLine($"Total: {totalTime:F2}ms");

            textBox1.Text = sb.ToString();

        }

        private void Form1_Load(object sender, EventArgs e)
        {
            startupPath = Application.StartupPath;

            string classer_path = "lable.txt";
            List<string> str = new List<string>();
            StreamReader sr = new StreamReader(classer_path);
            string line;
            while ((line = sr.ReadLine()) != null)
            {
                str.Add(line);
            }
            dicts = str.ToArray();

            image_path = "1.png";
            pictureBox1.Image = new Bitmap(image_path);
        }
    }
}

下载

源码下载

C# OpenVINO 直接读取百度Paddle模型实现物体检测( yolov3_darknet)

效果

项目

代码

下载

猜你喜欢