以下是使用Python、Node.js和Go创建基于YOLOv8的对象检测Web服务的简化版本。请注意，这些示例假设你已经安装了相应的库和环境。
- Python + FastAPI:
# Python 3.6 及以上
# detect.py
from fastapi import FastAPI
from starlette.responses import JSONResponse
import yolov8_inference
app = FastAPI()
@app.post("/detect/")
async def detect(image: bytes):
detections = yolov8_inference.detect_objects(image)
return JSONResponse(content=detections)- Node.js + Express:
// Node.js 12.x 及以上
// detect.js
const express = require('express');
const yolov8_inference = require('yolov8_inference');
const app = express();
const port = 3000;
app.use(express.json());
app.post('/detect/', async (req, res) => {
const image = req.body.image;
const detections = await yolov8_inference.detectObjects(image);
res.json(detections);
});
app.listen(port, () => {
console.log(`Server running on port ${port}`);
});- Go + Gin:
// Go 1.13 及以上
// detect.go
package main
import (
"github.com/gin-gonic/gin"
"yolov8_inference"
)
func main() {
router := gin.Default()
router.POST("/detect/", func(c *gin.Context) {
var imageBytes []byte
if err := c.ShouldBindJSON(&imageBytes); err != nil {
c.JSON(400, gin.H{"error": err.Error()})
return
}
detections := yolov8_inference.DetectObjects(imageBytes)
c.JSON(200, detections)
})
router.Run()
}请注意，上述代码假设yolov8_inference模块已经被安装并且包含DetectObjects函数，该函数接受图像字节并返回检测结果。在实际应用中，你需要替换为YOLOv8的正确API调用。