Object coordinates in Object analytics API

[embeddeddev4]

Is there any support for retrieving an object's coordinates when a detection event occurs for an object analytics scenario? I would like to see the x,y coordinates whenever an active trigger occurs in the Object in Area scenario. Right now it seems only the classType and object ID is being returned from my websocket connection to Axis events.

[github-actions[bot]]

This automatically generated reply acts as a friendly reminder.

Answers to your questions will most often come from the community, from developers like yourself. You will, from time to time, find that Axis employees answers some of the questions, but this is not a guarantee. Think of the discussion forum as a complement to other support channels, not a replacement to any of them.

If your question remains unanswered for a period of time, please revisit it to see whether it can be improved by following the guidelines listed in Axis support guidelines.

[rocky-ji]

Yes, Metadata will return all the data in the view, it does not contain any AOA area or line info.
AOA API will just output the event result, event triggered, object number.
To get what you want, you need get both data and check the coordinate of object and AOA area.

[embeddeddev4]

thanks for the helpful reply, so the metadata stream is how I get the data I want. I have been taking a look at the documentation and it doesn't support websocket connections correct? Only RTSP, ACAP sdk, and maybe MQTT from what I'm reading.

[vivekatoffice]

Hi @embeddeddev4 ,
I tried something previously that may help you:

Get the Polygon and scenario details using VAPIX and draw using OpenCV:

Device and AOA used:

AXIS Q1715 Block Camera
AXIS OS version 11.11.73
AXIS Object Analytics ( Version: 1.14.35)

import requests
from requests.auth import HTTPDigestAuth
import cv2
import numpy as np

# Replace with your actual device IP or hostname
device = "10.176.12.148"
user = "Vivek"
password = "Kumar"

# URL to request the configuration
url = f"http://{device}/local/objectanalytics/control.cgi"

# Body of the POST request
body = {
    "apiVersion": "1.2",
    "context": "my context",
    "method": "getConfiguration"
}

# Make the POST request using Digest Authentication
response = requests.post(url, json=body, auth=HTTPDigestAuth(user, password))

try:
    # Check if the request was successful
    response.raise_for_status()

    # Check if the response is empty or not
    if not response.text.strip():
        print("Empty response received from the server.")
        exit(1)

    # Print the response text to see what is being returned
    print("Response Text:", response.text)

    # Parse the response JSON
    config = response.json()

    # Extract relevant information from the JSON
    scenario = config["data"]["scenarios"][0]
    scenario_name = scenario["name"]
    include_area = scenario["triggers"][0]["vertices"]

    # RTSP URL (Use your actual RTSP URL here)
    rtsp_url = f"rtsp://{user}:{password}@{device}:554/axis-media/media.amp"

    # Open the RTSP stream
    cap = cv2.VideoCapture(rtsp_url)

    # Scaling function to map from normalized (-1.0 to 1.0) to image coordinates
    def scale_coordinates(coords, frame_width, frame_height):
        return [
            (
                int((x + 1.0) / 2.0 * frame_width),
                int((y + 1.0) / 2.0 * frame_height)
            )
            for x, y in coords
        ]

    # Process each frame
    while cap.isOpened():
        ret, frame = cap.read()
        if not ret:
            print("Failed to grab frame")
            break
        
        frame_height, frame_width = frame.shape[:2]

        # Draw include area
        include_pts = scale_coordinates(include_area, frame_width, frame_height)
        cv2.polylines(frame, [np.array(include_pts)], isClosed=True, color=(0, 255, 0), thickness=2)

        # Overlay the scenario name on the video
        cv2.putText(frame, scenario_name, (10, 30), cv2.FONT_HERSHEY_SIMPLEX, 1, (255, 255, 255), 2, cv2.LINE_AA)

        # Show the frame
        cv2.imshow('RTSP Stream with Overlay', frame)

        # Press 'q' to exit
        if cv2.waitKey(1) & 0xFF == ord('q'):
            break

    # Release the capture and close the window
    cap.release()
    cv2.destroyAllWindows()

except requests.exceptions.RequestException as e:
    print(f"Request failed: {e}")

Get the metadata (SVG data) using the Websocket interface:

Topic filter can be changed to get the AXIS Scene Metadata integration

{
        "eventFilterList": [
            {
                "topicFilter": "onvif:VideoAnalytics/axis:TrackEnded"
            },
            {
                "topicFilter": "onvif:VideoAnalytics/axis:TrackUpdated"
            }
        ]
  }

import json
import requests, re
import websockets
import asyncio
import hashlib
import base64
import uuid
import logging
logger = logging.getLogger('websockets')
logger.setLevel(logging.DEBUG)
logger.addHandler(logging.StreamHandler())

def calculate_digest_response(username, realm, password, uri, method, nonce, nc, cnonce):
    ha1 = hashlib.md5(f"{username}:{realm}:{password}".encode()).hexdigest()
    ha2 = hashlib.md5(f"{method}:{uri}".encode()).hexdigest()
    response = hashlib.md5(
        f"{ha1}:{nonce}:{nc}:{cnonce}:auth:{ha2}".encode()).hexdigest()
    return response


async def connect(nonce):
    username = 'Vivek'
    password = 'Kumar'
    realm = 'AXIS_B8A44FXXXXXX'  #B8A44F279CC0 is the the MAC address of the camera.
    uri = '/vapix/ws-data-stream?sources=events'
    method = 'GET'

    nc = '00000001'
    cnonce = str(uuid.uuid4())

    response = calculate_digest_response(
        username, realm, password, uri, method, nonce, nc, cnonce)
    print("Digest calculated response:", response, "\n")
    authorization_header = f'Digest username="{username}", realm="{realm}", nonce="{nonce}", uri="{uri}", response="{response}", qop=auth, nc={nc}, cnonce="{cnonce}"'
    print(" ---- Digest header ---- \n"+authorization_header+"\n")
    websocket_uri = "ws://10.176.12.148/vapix/ws-data-stream?sources=events"
    websocket_headers = {
        "Authorization": authorization_header
    }
    
    try:
        async with websockets.connect(websocket_uri, extra_headers=websocket_headers, user_agent_header=None) as websocket:
            print("WebSocket connection established.")

            # Prepare the JSON payload
            payload = {
                        "apiVersion": "1.0",
                        "method": "events:configure",
                        "params": {
                        "eventFilterList": [
                                              {
                                                "topicFilter": "tnsaxis:CameraApplicationPlatform/ObjectAnalytics/xinternal_data"
                                              }
                                            ]
                                  }
                      }
           

            # Convert the payload to JSON string
            payload_json = json.dumps(payload)

            # Send the JSON payload
            await websocket.send(payload_json)
            print("JSON payload sent.")

            # Continue with WebSocket communication
            while True:
                message = await websocket.recv()
                print(f"Received message: {message}")
    except Exception as e:
        print(e)        


def extract_value(header, key):
    match = re.search(fr'{key}="([^"]+)"', header)
    if match:
        return match.group(1)
    return None

def perform_http_request():
    username = 'Vivek'
    password = 'Kumar'
    url = 'http://10.176.12.148/axis-cgi/login.cgi'

    # Make a GET request to obtain the Digest authentication headers
    response = requests.get(url)
    #print(response.text)
    # Extract the required headers from the response
    nonce = response.headers.get('WWW-Authenticate')
    #print(nonce)
    match = re.search(r'nonce="([^"]+)"', nonce)
    if match:
        nonce = match.group(1)
        print("Nonce from response:", nonce,"\n")
    else:
        print("Nonce not found in the string.")
    response = requests.get(url, auth=requests.auth.HTTPDigestAuth(username, password))
    # Continue with the WebSocket connection using the obtained headers
    asyncio.run(connect(nonce))


if __name__ == "__main__":
    perform_http_request()

[embeddeddev4]

Thanks however this may be too cpu intensive for me as I just need to pass along more in depth details per detection like clothing color + coordinates which metadata provides. I have a question on the topic filters you gave.

            {
                "topicFilter": "onvif:VideoAnalytics/axis:TrackEnded"
            },
            {
                "topicFilter": "onvif:VideoAnalytics/axis:TrackUpdated"
            }

Are you sure these are the topicFilters for retrieving metadata via websocket events? I am not receiving any information when an object is detected using these filters on a camera with the latest firmware. I believe I will need to just implement metadata subscription using the sdk.

[vivekatoffice]

Hi @embeddeddev4 ,
Please try this API to enable the "selected_metadata_events" Feature Flag Service. I found this step after requesting the AXIS Scene Metadata MQTT Integration guide from here: AXIS Scene Metadata integration

The feature flag is available from AXIS OS 11.6.

1. Set the "selected_metadata_events" flag to true/false.

http://10.176.12.148/axis-cgi/featureflag.cgi
JSON input parameters

{ 
  "apiVersion": "1.0", 
  "context": "Client defined request ID", 
  "method": "set", 
  "params": { 
    "flagValues": { 
      "selected_metadata_events": true 
    } 
  } 
} 

2. Parse the JSON response.

{ 
    "apiVersion": "1.0", 
    "context": "Client defined request ID", 
    "method": "set", 
    "data": { 
        "result": "Success" 
    } 
} 

Output:

[mneser]

I only managed to get the detection data through via web socket after starting the feed with this request:
url = f'http://{ip}/local/objectanalytics/control.cgi'
payload = {
"apiVersion": "1.3",
"method": "startVisualConfirmation",
"params": {"scenario": 1, "transform": [1, 0, 0, 0, 1, 0, 0, 0, 1]}
}