Detect an Object

Detects an object within an uploaded image file. Make sure to load the model you're using for detection first!

Request

This endpoint expects a multipart form containing a file.

organization_idstringRequired

Your organization identifier

filefileRequired

An image file you'd like to run detection on

Response

200 - Detection successful

statusstring or null

A status code denoting success or failure.

detectionslist of any or null

The detections response includes 3 elements: the bbox array, a confidence value, and a class. The bbox array is an array of numbers describing the bounding box coordinates for your objects listed in order as [x1, y1, x2, y2]. The confidence score is a value between 0 and 1 rating how confident the object detection output is based on your model and the image provided. The class is the class name of the detected object.

1	import requests
2
3	url = "https://tytbrazyv9.execute-api.us-east-1.amazonaws.com/prod/inference/detect"
4
5	files = { "file": "open('<file1>', 'rb')" }
6	payload = { "organization_id": "organization_id" }
7	headers = {"x-api-key": "x-api-key"}
8
9	response = requests.post(url, data=payload, files=files, headers=headers)
10
11	print(response.json())

1	{
2	"status": "success",
3	"detections": [
4	{
5	"bbox": [
6	100,
7	150,
8	300,
9	400
10	],
11	"confidence": 0.95,
12	"class": "object"
13	}
14	]
15	}

Headers

Request

Response

Errors