Documentation
test_url : https://complete-goose-helping.ngrok-free.app
Method:
- POST
URI:
- {test_url}/classify_semaphore
Headers:
- Content-Type: multipart/form-data
Request Body:
imageasfile, must be a valid image, large size
Response:
200
- Detected image confidence > 0.5
{
"success": true,
"message": {
"confidence": 1.0,
"predicted_class": "B"
}
}
- Detected image confidence < 0.5
{
"success": true,
"message": {
"confidence": 0.2,
"predicted_class": "Not a Semaphore"
},
}
400
- Invalid file format
{
"success": false,
"message": "Invalid file format. Please upload a JPG, JPEG, or PNG image."
}
*notes: for better classification, image should be taken potrait in camera or image size ratio should be 3:4