请求一个yoloV8的使用案例，我已经折腾了很久，但是不知道解决方案，麻烦参考我代码，请求指引

[cplasfwst]

`package main

import (
"fmt"
"image"
"image/color"
"log"

"gocv.io/x/gocv"

)

var yolo_classes = []string{
"xiaohongshu",
}

func main() {

imgg := gocv.IMRead(`screenshot.png`, gocv.IMReadColor)
if imgg.Empty() {
	fmt.Println(`“无法加载图像文件”`)
	return
}
// 创建一个用于存储转换结果的 Mat 对象
//var converted gocv.Mat
// 调整图像尺寸到 640x640
resizedImg := gocv.NewMat()
gocv.Resize(imgg, &resizedImg, image.Point{X: 640, Y: 640}, 0, 0, gocv.InterpolationDefault)
//gocv.Resize(img, &resized, image.Point{X: 416, Y: 416}, 0, 0, gocv.InterpolationDefault)
//resizedImg.ConvertTo(&resizedImg, gocv.MatTypeCV32F, 1.0/255.0)
// 将图像数据转换为浮点数类型
//resizedImg.ConvertTo(&resizedImg, gocv.MatTypeCV32F)

// 归一化处理
gocv.ConvertScaleAbs(resizedImg, &resizedImg, 1.0/255.0, 0)
defer resizedImg.Close()

fmt.Println("resizedImg dimensions:", resizedImg.Rows(), resizedImg.Cols(), resizedImg.Channels())
// 检查图像信息
fmt.Println("resizedImg dimensions:", resizedImg.Rows(), resizedImg.Cols(), resizedImg.Channels())
// 加载 YOLOv8 模型
net := gocv.ReadNetFromONNX("best.onnx")
if net.Empty() {
	log.Fatal("Error loading YOLOv8 model")
}
defer net.Close()

//打印所有层的名称
//layerNames := net.GetLayerNames()
//for i, name := range layerNames {
//	fmt.Printf("模型的层名：Layer %d: %s\n", i, name)
//}

// 创建 blob 时，确保输入形状、均值、缩放因子等符合模型要求
blob := gocv.BlobFromImage(resizedImg, 1.0/255.0, image.Pt(640, 640), gocv.NewScalar(0, 0, 0, 0), true, false)
net.SetInput(blob, "images")
defer blob.Close()

// 打印 blob 的具体信息
fmt.Println("Blob dimensions:", blob.Size())
fmt.Println("Blob type:", blob.Type())

// 执行前向传播
detections := net.Forward("output0")
if detections.Empty() {
	log.Fatal("Detection output is empty")
}
defer detections.Close()

// 打印 detections 的形状
fmt.Println("Detections shape:", detections.Size())

fmt.Println("size:", detections.Size())
// 打印检测结果的形状以进行调试
fmt.Printf("Detection shape: %d x %d\n", detections.Rows(), detections.Cols())

// 打印 detections 的维度
shape := detections.Size()
fmt.Printf("Detection shape: %d x %d x %d\n", shape[0], shape[1], shape[2])

// 解析检测结果
confidenceThreshold := float32(0.5)
classesThreshold := float32(0.5)
// 按模型的输出格式调整解析
for i := 0; i < shape[2]; i++ { // 遍历每个检测框
	confidence := detections.GetFloatAt(0, 4+i*5) // 可能需要调整索引
	if confidence > confidenceThreshold {
		// 解析类别置信度
		var classID int
		classConfidence := float32(-1.0)
		for c := 0; c < len(yolo_classes); c++ {
			classConf := detections.GetFloatAt(0, 5+c+i*5) // 可能需要调整索引
			if classConf > classConfidence {
				classConfidence = classConf
				classID = c
			}
		}
		if classConfidence > classesThreshold {
			// 获取边界框坐标
			x := detections.GetFloatAt(0, i*5) * float32(resizedImg.Cols())
			y := detections.GetFloatAt(0, i*5+1) * float32(resizedImg.Rows())
			width := detections.GetFloatAt(0, i*5+2) * float32(resizedImg.Cols())
			height := detections.GetFloatAt(0, i*5+3) * float32(resizedImg.Rows())

			// 获取类别名称
			className := "unknown"
			if classID < len(yolo_classes) {
				className = yolo_classes[classID]
			}

			// 打印边界框坐标、置信度和类别
			fmt.Printf("Detected %s at (%.0f, %.0f, %.0f, %.0f) with confidence %.2f\n", className, x, y, width, height, confidence)

			// 在图像上绘制检测结果
			gocv.Rectangle(&imgg, image.Rect(int(x-width/2), int(y-height/2), int(x+width/2), int(y+height/2)), color.RGBA{255, 0, 0, 0}, 2)
			gocv.PutText(&imgg, fmt.Sprintf("%s: %.2f", className, confidence), image.Pt(int(x-width/2), int(y-height/2)-10), gocv.FontHersheyPlain, 1, color.RGBA{255, 0, 0, 0}, 2)
		}
	}
}

// 保存绘制后的图像
if ok := gocv.IMWrite("annotated_screenshot.png", imgg); !ok {
	log.Fatal("Failed to write annotated screenshot")
}

// 显示结果
window := gocv.NewWindow("YOLOv8 Detections")
window.IMShow(imgg)
window.WaitKey(0)
window.Close()

}
`

请求一个完整的yolo案例，有谁能帮帮我

[swdee]

An example/demo has been created here https://github.com/hybridgroup/gocv/blob/release/cmd/yolo-detection/main.go