Predict water-hyacinth coverage and bounding boxes for one image

POST

/api/v1/water-hyacinth/predict

Last modified:2026-06-22 20:10:30

Maintainer:Not configured

Runs the notebook-derived classifier and detector on one uploaded JPG/PNG image. The response includes YOLO bounding boxes that mobile or web clients can draw as overlays. The coverage percentage and risk level are computed with the same formula used in backend/reference/single_image_prediction.py.

Request

Body Params multipart/form-dataRequired

Responses

🟢200

application/json

Successful Response

Bodyapplication/json

🟠422

Request Example

Shell

JavaScript

Java

Swift

curl --location 'http://13.60.162.222:3000/api/v1/water-hyacinth/predict' \
--form 'file=@""'

Response Example

200 - Example 1

{
    "classification": "string",
    "classification_confidence": 0,
    "detected_regions": 0,
    "coverage_percent": 0,
    "detection_confidence": 0,
    "risk_level": "string",
    "image_width": 0,
    "image_height": 0,
    "boxes": [
        {
            "x1": 0,
            "y1": 0,
            "x2": 0,
            "y2": 0,
            "confidence": 0,
            "width": 0,
            "height": 0
        }
    ]
}

Modified at 2026-06-22 20:10:30

Describe the mounted API base path

Predict water-hyacinth coverage and bounding boxes for multiple images