1. Project Overview

Sithong AI Video Surveillance System - Completely Free & Code Open-source

Focused on Visual AI Algorithm Customization, 1000+ AI Scene Landing Experience

Produce a New Algorithm in as Little as 8 Hours

AI Video Surveillance System, Completely Free & Code Open-source. Supports Localized Deployment with Docker One-click Installation

2. Open Source License

Licensed under GPL-3.0

3. Screenshots

(Commercial Version - Partial Screenshot)

Training Model List

Label List

Image Annotation

4. Feature Overview

Project Overview

This product is a multi-functional AI real-time monitoring system, supporting multiple device mixed access, including: drones, dash cams, cash registers, surveillance cameras, pads, mobile phones, and computer cameras. The system has high flexibility, and can provide real-time warnings for monitoring targets and ranges. In addition, the system has a built-in low-code algorithm model training platform, which simplifies the algorithm development process. Users can also conveniently search for faces, objects, labels, license plates, and voice conversations in monitoring video through text or image search.

Product Functions

5. Tech Stack

Technical Stack Introduction

• Video Storage: Minio
• Message Middleware: RabbitMQ
• Database: MySQL
• Video Search Index: Elasticsearch
• Video Algorithm: YoloV8
• Warning Message Push: Websocket
• Video Stream Transcoding: ffmpeg
• System Cache: Redis

Machine Configuration

Open-source Models (20 free models)

Model Name	Version
Wearing Safety Helmet	v1.0
Human Fall Detection	v1.0
Large Truck Monitoring	v1.0
Mask Recognition	v1.0
Glove Recognition	v1.0
Smoking Monitoring	v1.0
Identifying Phone Calls	v1.0
Fighting Detection	v1.0
Wheelchair Detection	v1.0
Stroller Detection	v1.0
Traffic Accident Detection	v1.0
Coal Tank Detection	v1.0
Dump Truck Detection	v1.0
Drone Detection	v1.0
Water Gauge Scale	v1.0
Fishing Detection	v1.0
Ship Detection	v1.0
Cat and Dog Detection	v1.0
Small Boat Detection	v1.0

Docker Installation

1. Machine Minimum Configuration Requirements: CPU intel i5, Memory 8G, Disk 80G.

2. Users can get the docker image and install it in about 10-15 minutes on either Windows or Linux operating systems by running a single command.

3. Open the browser and enter the access address: http://your_IP_address:35300/vis/, you can see the system interface after local installation is successful.