C#实现控制台传参调用YoloV5模型进行人体识别

合集 - 机器视觉-Yolo(8)

1.详细Anaconda+Pycharm配置YoloV5环境笔记2021-09-032.Pytorch学习笔记-（二）搭建Pytorch环境（Pytorch2.1+CUDA12.1+Anaconda3_2023+Pycharm2023）2023-10-093.yolov5-OSError: [WinError 1455] 页面文件太小，无法完成操作。 Error loading "C:\Anaconda3\lib\site-packages\torch\lib\caffe2_detectron_ops_gpu.dll"2021-09-074.YoloV5_RuntimeError: CUDA out of memory. Tried to allocate 100.00 MiB (GPU 0; 2.00 GiB total capacity; 1.15 GiB already allocated; 0 bytes free; 1.19 GiB reserved in total by PyTorch)2021-09-075.C#调用YoloV5与YoloV8示例2024-05-11

6.C#实现控制台传参调用YoloV5模型进行人体识别2024-07-08

7.ML.NET-API（一）_常用代码片段2024-07-128.ML.NET-模型生成器工具（二）-对象检测教程2024-07-12

收起

一、项目地址

  https://gitee.com/qq28069933146_admin/yoloconsole_body

二、代码解析

1、获取输入参数

//static void Main(string[] args)方法中运行如下代码：
string toBeTestedImgPathStr = AppContext.BaseDirectory + "\\test.jpg";
if (args.Length > 0)
{
    toBeTestedImgPathStr = args[0];
}

2、C#加载YoloV5的onnx模型

YoloScorer<YoloCocoP5Model> scorer = new YoloScorer<YoloCocoP5Model>(modelPath);  // 选择模型

3、使用YoloV5模型进行预测

List<YoloPrediction> predictions = scorer.Predict(image);  // 预测

4、自定义人体信息类

  该类用于在控制台打印人体位置信息

    /// <summary>
    /// 人体信息
    /// </summary>
    public class BodyInfo
    {
        public int Left { get; set; } = 0;
        public int Top { get; set; } = 0;
        public int Right { get; set; } = 0;
        public int Bottom { get; set; } = 0;

        /// <summary>
        /// 概率
        /// </summary>
        public double Score { get; set; } = 0;

        public BodyInfo(int left, int top, int right, int bottom, double score)
        {
            Left = left;
            Top = top;
            Right = right;
            Bottom = bottom;
            Score = score;
        }
    }

5、用BodyInfo保存人体信息并在图片上标注人体位置

  // 保存人体位置信息
  List<BodyInfo> bodyInfos = new List<BodyInfo>();

  // 标注图片
  foreach (YoloPrediction prediction in predictions)
  {
      if (prediction.Label.Name == "person")  // prediction.Label.Id == "1"
      {
          double score = Math.Round(prediction.Score, 2);                               // 概率（保留两位小数）
          var (x, y) = (prediction.Rectangle.Left - 3, prediction.Rectangle.Top - 23);  // 标注信息的位置信息

          // 标说明
          image.Mutate(a => a.DrawText($"{prediction.Label.Name} ({score})",
                          font, prediction.Label.Color, new PointF(x, y)));

          // 框位置
          image.Mutate(a => a.DrawPolygon(prediction.Label.Color, 1,
              new PointF(prediction.Rectangle.Left, prediction.Rectangle.Top),
              new PointF(prediction.Rectangle.Right, prediction.Rectangle.Top),
              new PointF(prediction.Rectangle.Right, prediction.Rectangle.Bottom),
              new PointF(prediction.Rectangle.Left, prediction.Rectangle.Bottom)
              ));

          bodyInfos.Add(new BodyInfo((int)prediction.Rectangle.Left, (int)prediction.Rectangle.Top, (int)prediction.Rectangle.Right, (int)prediction.Rectangle.Bottom, score));
      }
  }

6、保存图片

image.SaveAsync(toBeTestedImgNewNameStr);

7、控制台输出结果(json字符串格式)

string resultStr = JsonConvert.SerializeObject(bodyInfos);

Console.WriteLine(resultStr);

8、运行效果

  控制台调用，输入参数即可（如：YoloConsole.exe -E：）；效果如下：