Smart Video Analytics AI (YOLOv8)

Turn any video footage into actionable intelligence. Detect weapons, analyze traffic, count crowds, and monitor retail behavior using advanced AI computer vision.

---

What is Smart Video Analytics AI?

Smart Video Analytics AI is a powerful AI tool that processes video files to detect specific objects, behaviors, and security threats. Built on the state-of-the-art YOLOv8 architecture, it scans video footage frame-by-frame, identifies relevant objects (like weapons, vehicles, or unattended bags), and generates a detailed report with timestamps and photographic evidence.

Unlike generic object detectors, this Actor features 15+ specialized Analysis Modes. Whether you are a security professional looking for threats, a city planner analyzing traffic, or a business owner tracking store occupancy, this tool filters out the noise and only reports what matters to you.

Why use this Actor?

Video analysis is traditionally expensive and time-consuming. Humans cannot watch 24/7 feeds effectively, and enterprise AI solutions often cost thousands of dollars.

Smart Video Analytics AI solves this by:

• Automating Surveillance: Process hours of footage in minutes.
• Reducing False Positives: Select specific modes (e.g., "Fire Safety") to only look for relevant object combinations (e.g., Cars blocking Fire Hydrants).
• Smart Evidence: It automatically captures and hosts a screenshot (JPG) for every detection, giving you instant visual proof without downloading the whole video.

Main Features

• 15 Specialized Modes: From Threat Detection to Retail Analytics and Wildlife Monitoring.
• Multi-Mode Analysis: Run multiple checks simultaneously (e.g., count cars AND detect weapons in the same pass).
• Configurable Speed: Adjust the frameStep to balance between high-speed scanning and frame-perfect accuracy.
• JSON & Excel Reports: Get structured data compatible with Tableau, PowerBI, or your own dashboard.

Analysis Modes

Select one or multiple modes in the input to tailor the AI's focus:

Mode	Detects	Best For
🚨 Threat Detection	Persons, Knives, Bats, Scissors	Security teams, Crime analysis
🚗 Traffic Surveillance	Cars, Trucks, Buses, Bikes	City planning, Road usage stats
⛽ Petrol Garage	Vehicles, Persons, Trucks	Fuel station logistics
🛍️ Retail Analytics	Persons, Handbags, Backpacks	Store heatmaps, Loss prevention
🎒 Abandoned Object	Suitcases, Backpacks	Airports, Train stations
🔥 Fire Safety	Hydrants, Vehicles	Detecting blocked emergency access
👥 Crowd Counting	Persons	Occupancy limits, Queue management
🚜 Heavy Machinery	Trucks	Construction site monitoring
👤 Face Logging	Persons (Body tracking)	Timestamping entry/exit
🐶 Pet Behavior	Cats, Dogs, Furniture	Home monitoring applications

How to analyze a video

This Actor is designed to be simple. [cite_start]You do not need coding skills to use it[cite: 6].

1. Get your Video URL: Ensure you have a direct link to a video file (.mp4, .avi, .mkv). This can be from cloud storage (Google Cloud, S3, Azure) or a public server.
2. Select Analysis Mode: Choose the logic you want to apply (e.g., "Threat Detection"). You can select multiple.
3. Set Frame Step:
   • Use 30 (default) to check 1 frame per second (assuming 30fps video).
   • Use 1 for frame-perfect accuracy (slower).
4. Run: Click Start. The Actor will output a list of detections as they happen.

Input Example

[cite_start]You can configure the actor using a simple JSON object or the visual interface[cite: 4].

{
    "videoUrl": "[https://example.com/security_footage.mp4](https://example.com/security_footage.mp4)",
    "analysisMode": ["threat_detection", "abandoned_object"],
    "frameStep": 15,
    "confidenceThreshold": 0.5
}

Output Example

[cite_start]The Actor produces a dataset where every item represents a specific frame where a detection occurred[cite: 5].

{
    "timestamp_seconds": 12.5,
    "frame_index": 375,
    "triggered_modes": ["threat_detection"],
    "objects_detected": ["person (0.92)", "knife (0.88)"],
    "object_count": 2,
    "screenshot_url": "[https://api.apify.com/v2/key-value-stores/](https://api.apify.com/v2/key-value-stores/)..."
}

Pricing & Costs

This Actor uses deep learning models which require CPU power. [cite_start]Pricing is based on the duration of the video processed and the frame step you select[cite: 3].

• Estimated Cost: ~$0.05 per 10 minutes of video analysis (at 30 frame step).
• Optimization Tip: To save costs on long static videos (like overnight security footage), increase the frameStep to 60 or 100. This reduces the number of AI inferences required.

Legal & Licensing

Open Source Compliance (AGPL-3.0)

This Actor utilizes YOLOv8 by Ultralytics. In compliance with the AGPL-3.0 License, the source code for this Actor is publicly available.

• License: You are free to use this for commercial purposes, provided that if you modify and redistribute this software as a service, you must also share your source code.

Data Privacy & GDPR

We do not store your video files. Videos are processed in volatile memory and discarded immediately after analysis. Screenshots are stored in your private Apify Key-Value store and are accessible only by you (unless you share the public link). [cite_start]Do not use this tool to infringe on the privacy rights of others[cite: 7].

Support & Feedback

If you encounter any issues or have a request for a new "Mode" (e.g., detecting specific industrial equipment), please reach out!

• Report Bugs: Use the "Issues" tab on this page.
• Custom Solutions: We can train custom models (e.g., detecting guns, specific uniforms, or brand logos) for enterprise clients. Contact us for details.

undefined