Main Workflow Guide

Quick Connection

With only an RTSP address, you can register a camera in EVA and configure AI analysis immediately.

EVA attempts connection as soon as you enter the IP camera RTSP streaming URL. If successful, it retrieves live video and starts AI analysis without complex setup.

You can also register the same RTSP address multiple times as separate logical cameras, and apply different scenarios to each for multi-angle analysis.

Registration Steps

1. Click Add New Camera on the EVA dashboard.

2. Enter only camera name and RTSP URL.

   • Example: rtsp://username:password@192.168.1.100:554/stream

3. A connection test runs automatically. If successful, a still image is shown.

4. Registration complete. EVA fetches frames at the configured interval for AI analysis.

5. Check overall real-time status such as connection health and event counts on the dashboard.

⚠️ If the RTSP URL is incorrect or the network is unstable, connection can fail.

---

Camera Settings

For registered cameras, you can edit key settings including name and RTSP URL.

You can also register multiple cameras with the same RTSP and apply different detection scenarios.

How to Edit

1. Click a camera in the dashboard list to open its detail page.

2. Click the pencil icon (Edit) in the top-right corner.

3. You can edit:

   • Camera name
   • RTSP URL
   • Detection scenario
   • Detection targets
   • Alert channel settings
   • AI inference interval, etc.

4. Click Save to apply immediately.

Webhook Settings

When configured risk situations are detected, EVA sends alerts immediately for fast response.

EVA supports alerts to Teams, Slack, and n8n (Webhook).

1. Create Webhook URL in Teams/Slack.
2. Register in EVA from alert settings. Multiple webhooks can be configured by role/priority.
3. Received alert payload includes:
   • Detection message
   • EVA-generated detailed description
   • Related image snapshot

n8n tip

• Use n8n Webhook Trigger to extend flows like ticket creation, notifications, and automated actions.

For webhook creation in Teams/Slack, refer to the advanced settings documentation.

How to Delete

1. Click trash icon (Delete) on the detail page.
2. Select Delete in the confirmation dialog.
3. All camera data (analysis history, settings, etc.) is permanently removed.
4. Verify that the camera is removed from the list.

In streaming view, object detection runs continuously, so you can verify whether current targets are detected as intended.

However, scenario-based reasoning and alerting are enabled only after monitoring starts.

ℹ️ For Detection Scenario and Detection Target setup, see the relevant scenario and target/ROI guide pages.

---

Supported Address Formats and Camera Requirements

EVA primarily supports RTSP. For stable real-time streaming, meet the following requirements.

Supported Address Formats

Protocol	Example Address
RTSP (recommended)	rtsp://[user]:[pass]@[IP]:[port]/[path]
RTMP	rtmp://[IP]:[port]/[app]/[stream]
HTTPS (MJPEG)	https://[IP]/video.mjpg

Required Camera Specs

Item	Requirement
Resolution	Recommended 720p+ (minimum 640x480)
Frame Rate	5 fps+ (linked to AI inference interval)
Codec	H.264 / H.265 (limited MJPEG support)
Network	Stable wired or Wi-Fi (public IP/port-forwarding if needed)
Auth	ID/PW can be included in RTSP URL

Connection Tips

• Use internal IP on local networks
• For external access, configure DDNS + port forwarding or VPN
• Open RTSP port (default 554) in firewall

Maximum Number of Cameras

Maximum connectable cameras depend on EVA server hardware (CPU, GPU, RAM, network bandwidth).

Higher-performance servers can stably process more cameras concurrently.

---

Additional Features

• Polygon ROI support: Define ROI with polygons as well as rectangles.
• Brightness settings UI: Adjust brightness in camera detail modal with Auto/Manual/Unset mode. (slider -50 to +50)
• Chat-based brightness change disabled: Brightness can be changed only in settings UI.
• User-role feedback input: User role can also submit true/false feedback.