How to convert the following Python code into C # code to make it the same code?

[bbhxwl]

import torch
import torchvision.transforms as transforms
from torchvision.models.detection import ssd300_vgg16
from PIL import Image
import numpy as np


def preprocess_image(image_path):
    """
    预处理输入图像，将其调整为模型需要的输入格式。
    """
    # 定义图像转换
    transform = transforms.Compose([
        transforms.Resize((300, 300)),  # 调整图像大小
        transforms.ToTensor(),  # 转换为张量
        transforms.Normalize(mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225])  # 标准化
    ])

    # 打开图像并应用转换
    image = Image.open(image_path).convert('RGB')
    image_tensor = transform(image).unsqueeze(0)  # 添加批次维度

    return image_tensor

def run_inference(image_tensor, threshold=0.5):
    """
    使用 SSD 模型进行推理，并根据置信度分数过滤结果。
    """
    # 加载预训练的 SSD 模型
    model = ssd300_vgg16(pretrained=True)
    model.eval()  # 切换模型到评估模式

    with torch.no_grad():
        # 进行推理
        predictions = model(image_tensor)

    # 提取检测结果
    boxes = predictions[0]['boxes'].numpy()
    labels = predictions[0]['labels'].numpy()
    scores = predictions[0]['scores'].numpy()

    # 根据置信度分数过滤结果
    filtered_boxes = boxes[scores > threshold]
    filtered_labels = labels[scores > threshold]
    filtered_scores = scores[scores > threshold]

    return filtered_boxes, filtered_labels, filtered_scores

def print_results(boxes, labels, scores):
    """
    打印检测结果。
    """
    print("检测结果：")
    for i in range(len(labels)):
        print(f"Box: {boxes[i]}")
        print(f"Label: {labels[i]}")
        print(f"Score: {scores[i]:.2f}\n")

image_path = "/Users/xuzhibin/Downloads/6ee927a0d4f2c9862a918798de175f5.jpg"  
image_tensor = preprocess_image(image_path)
boxes, labels, scores = run_inference(image_tensor)
print_results(boxes, labels, scores)

using ConsoleApp10;
using Microsoft.ML.OnnxRuntime;
using Microsoft.ML.OnnxRuntime.Tensors;

string modelPath = "/Users/xuzhibin/Desktop/python/myScript/ssd300_vgg16.onnx";
using var session = new InferenceSession(modelPath);
var inputData=test.PreprocessImage("/Users/xuzhibin/Downloads/6ee927a0d4f2c9862a918798de175f5.jpg");
// 创建张量
var inputTensor = new DenseTensor<float>(inputData, new[] { 1, 3, 300, 300 });

// 设置输入
var inputs = new List<NamedOnnxValue>
{
    NamedOnnxValue.CreateFromTensor("input", inputTensor)
};

// 运行推理
using IDisposableReadOnlyCollection<DisposableNamedOnnxValue> results = session.Run(inputs);

// 获取输出
var boxes = results.FirstOrDefault(x => x.Name == "boxes").AsEnumerable<float>().ToArray();
var labels = results.FirstOrDefault(x => x.Name == "labels").AsEnumerable<long>().ToArray();
var scores = results.FirstOrDefault(x => x.Name == "scores").AsEnumerable<float>().ToArray();

// 输出前几个检测结果
Console.WriteLine("检测结果：");
for (int i = 0; i < Math.Min(5, labels.Length); i++)
{
    Console.WriteLine($"Box: {boxes[i * 4 + 0]}, {boxes[i * 4 + 1]}, {boxes[i * 4 + 2]}, {boxes[i * 4 + 3]}");
    Console.WriteLine($"Label: {labels[i]}");
    Console.WriteLine($"Score: {scores[i]}");
}

using System.Drawing;
using System.Drawing.Imaging;
using SkiaSharp;

namespace ConsoleApp10;

public class test
{
    public static float[] PreprocessImage(string imagePath)
    {
        // 读取图像为 SKBitmap
        using var inputStream = File.OpenRead(imagePath);
        using var originalBitmap = SKBitmap.Decode(inputStream);
        // 创建新的 Bitmap，调整到模型需要的大小
        using var resizedBitmap = originalBitmap.Resize(new SKImageInfo(300, 300), SKFilterQuality.High);
        // 创建数据数组
        var inputData = new float[1 * 3 * 300 * 300];
        int index = 0;

        // 遍历每个像素，将颜色值归一化到 [0, 1] 范围
        for (int y = 0; y < resizedBitmap.Height; y++)
        {
            for (int x = 0; x < resizedBitmap.Width; x++)
            {
                var pixel = resizedBitmap.GetPixel(x, y);
                inputData[index++] = pixel.Red / 255.0f;
                inputData[index++] = pixel.Green / 255.0f;
                inputData[index++] = pixel.Blue / 255.0f;
            }
        }

        return inputData;
    }
}