Images

Overview

Most Viewtron camera events include snapshot images — a source image (full camera view at the time of detection) and optionally a target image (cropped area of the detected object). Images are JPEG format, embedded as base64 in the XML payload.

Image Methods

Available on all event objects (LPR, Face, Intrusion, Counting):

Method	Returns	Description
source_image_exists()	bool	Whether a source (overview) image is attached
target_image_exists()	bool	Whether a target (cropped) image is attached
images_exist()	bool	Whether any images are attached
get_source_image()	str or None	Base64-encoded source image
get_target_image()	str or None	Base64-encoded target image
get_source_image_bytes()	bytes or None	Decoded JPEG bytes for source image
get_target_image_bytes()	bytes or None	Decoded JPEG bytes for target image

Source vs Target Images

Image Type	What It Contains	Typical Use
Source image	Full camera view at the moment of detection	Scene context, recording, human review
Target image	Cropped region of the detected object	Face crops, plate crops, object identification

When Images Are Included

Event Type	Source Image	Target Image
LPR	Yes — full scene	Yes — cropped plate
Face Detection	Yes — full scene	Yes — cropped face
Intrusion	Yes — full scene	Sometimes
Counting	Sometimes	Rarely
Traject	No	No

Image availability depends on camera model and configuration. Always check source_image_exists() / target_image_exists() before accessing.

Code Examples

Saving Images to Disk

def handle_event(event, client_ip):
    timestamp = event.get_time_stamp()
    category = event.category

    if event.source_image_exists():
        with open(f"{category}_{timestamp}_source.jpg", "wb") as f:
            f.write(event.get_source_image_bytes())

    if event.target_image_exists():
        with open(f"{category}_{timestamp}_target.jpg", "wb") as f:
            f.write(event.get_target_image_bytes())

Sending via MQTT

import paho.mqtt.client as mqtt

mqtt_client = mqtt.Client()
mqtt_client.connect("localhost", 1883)

def handle_event(event, client_ip):
    if event.category == "lpr" and event.source_image_exists():
        # Publish raw JPEG bytes
        mqtt_client.publish(
            "camera/lpr/image",
            event.get_source_image_bytes()
        )

Base64 for Web/API

import json

def handle_event(event, client_ip):
    if event.category == "face" and event.target_image_exists():
        # Base64 string ready for JSON or HTML <img> tags
        face_b64 = event.get_target_image()

        payload = json.dumps({
            "camera": event.get_ip_cam(),
            "face_image": face_b64
        })
        # Send to your API, embed in HTML as data:image/jpeg;base64,...

Checking Before Accessing

def handle_event(event, client_ip):
    if not event.images_exist():
        print(f"No images in {event.category} event")
        return

    if event.source_image_exists():
        source_bytes = event.get_source_image_bytes()
        print(f"Source image: {len(source_bytes)} bytes")

    if event.target_image_exists():
        target_bytes = event.get_target_image_bytes()
        print(f"Target image: {len(target_bytes)} bytes")

Image Sizes

Typical image sizes vary by camera resolution and model:

• Source images: 50KB-500KB (full resolution camera snapshot)
• Target images: 5KB-50KB (cropped region)

These are embedded in the XML payload, so events with images are significantly larger than events without.

Related Pages

• LPR Events — plate number with plate crop images
• Face Detection Events — face detection with face crop images
• ViewtronEvent — common event properties