Propped door detection API
Detect doors held open by wedges, boxes, chairs or other objects.
What it detects
Check access doors for objects keeping them open.
- Detection
- A door held open by an object: wedge, box, cone, chair, trash can, rock, or tape over the latch.
- Outside the definition
- No door propped by an object. A door held by a person, open while someone walks through, or an entrance standing open with nothing propping it.
Send a camera image.
Set KIRKWOOD_API_KEY on your server. Save a frame as frame.jpg.
import base64
import json
import os
import urllib.request
with open("frame.jpg", "rb") as frame:
image = base64.b64encode(frame.read()).decode()
body = json.loads('''{
"images": [
"<BASE64_JPEG>"
],
"tags": [
"door_propped_open"
],
"detections": {
"return_boxes": true
}
}''')
body["images"] = [image]
request = urllib.request.Request(
"https://app.kirkwood.ai/v1/analyze",
data=json.dumps(body).encode(),
headers={
"Authorization": "Bearer " + os.environ["KIRKWOOD_API_KEY"],
"Content-Type": "application/json",
},
method="POST",
)
with urllib.request.urlopen(request, timeout=120) as response:
print(json.dumps(json.load(response), indent=2))Read the result
The door_propped_open tag returns an answer and p_yes. It can be requested together with blocked-exit detection.
Camera setup
Show the bottom of the door and the latch area where possible. Include doors held by a person and doors opening normally in your tests.
Try it on your footage.
Run this tag through the API using the Python example, or request a live demo to try it with us.
Approved account required. Analysis uses credit. Images are retained for annotation and training.
Request a live demoBefore you integrate
An open door alone is not enough. The definition requires an object holding it open, such as a wedge or tape over a latch.
Does any open door count?
No. A person holding a door, someone walking through it or an entrance standing open without an object propping it is outside the positive definition.