Object Detection API

Request

• Method: POST
• URL: /api/qwen-open-ai
• Headers:
  • Content-Type: application/json
• Body:

  {
    "imageBase64": "base64-encoded image data",
    "prompt": "optional custom instruction, uses default if not provided"
  }

Response

• Success Response:

  {
    "bbox_2d": [[x1, y1, x2, y2]],
    "label": ["detected object labels"],
    "description": "object description"
  }

• Error Response:

  {
    "error": "error message",
    "code": error status code
  }

Error Codes

• 401: User not logged in
• 403: Insufficient credits
• 400: Image data is empty
• 500: Server internal error

Example

// Example request
const response = await fetch('/api/qwen-open-ai', {
  method: 'POST',
  headers: {
    'Content-Type': 'application/json'
  },
  body: JSON.stringify({
    imageBase64: "data:image/jpeg;base64,/9j/4AAQSkZJRg...",
    prompt: "Detect all vehicle positions in the image"
  })
});

// Example response
{
  "bbox_2d": [[100, 150, 300, 350], [400, 200, 600, 400]],
  "label": ["car", "truck"],
  "description": "The image contains two motor vehicles"
}

Notes

1. User login and sufficient credits are required
2. Image base64 data should not exceed 10MB
3. Default prompt: "Outline the positions of traffic accidents in the image and output all coordinates in JSON format"
4. The returned bbox_2d format is a 2D array of [x1,y1,x2,y2]

Code Examples

curl -X POST "http://segvision.satxspace.org/api/qwen-open-ai" \
-H "Content-Type: application/json" \
-d '{
  "imageBase64": "data:image/jpeg;base64,your-image-base64-data",
  "prompt": "Detect traffic accidents in the image"
}'