Skip to content

OwnDing/gb28181

BranchesTags

Repository files navigation

GB28181 Video Surveillance Platform / GB28181 视频监控平台

A modern, high-performance video surveillance management platform based on the GB/T 28181-2016 standard. Supports device auto-registration, real-time preview (H.264/H.265), cloud storage management, and AI-powered object detection.

基于 GB/T 28181-2016 标准构建的现代化高性能视频监控管理平台。支持设备自动注册、由 ZLMediaKit 驱动的实时预览（H.264/H.265）、云端存储管理以及基于 AI 的目标检测。

🏗 System Architecture / 系统架构

The system consists of four core components: 系统主要由以下四个核心组件构成：

  1. Backend (Java/Spring Boot): Handles GB28181 SIP signaling, device management, user authentication, and business logic.
    • 后端 (Java/Spring Boot)：处理 GB28181 SIP 信令、设备管理、用户认证及业务逻辑。
  2. Frontend (React/Vite): A modern web interface for device management and video playback.
    • 前端 (React/Vite)：现代化的 Web 界面，用于设备管理和视频播放。
  3. Media Server (ZLMediaKit): High-performance streaming server supporting RTSP/RTMP/HTTP-FLV/HLS/WebRTC.
    • 流媒体服务器 (ZLMediaKit)：高性能流媒体服务，支持 RTSP/RTMP/HTTP-FLV/HLS/WebRTC 等多种协议。
  4. AI Service (Python/YOLOv8): Real-time object detection (e.g., person detection) on video streams.
    • AI 服务 (Python/YOLOv8)：基于 YOLOv8 的实时视频流目标检测（如人形检测）。
graph TD
    Camera["IPC / NVR (GB28181)"] <-->|SIP| Backend("Java Backend")
    Camera -->|RTP/UDP/TCP| ZLM("ZLMediaKit")
    
    User["Web Browser"] <-->|HTTP/WebSocket| Frontend("React Frontend")
    Frontend <-->|API| Backend
    Frontend <-->|WebRTC/FLV/HLS| ZLM
    
    Backend <-->|API| ZLM
    Backend <-->|API| AI("AI Service")
    
    AI -->|RTSP| ZLM
    AI -->|Alarm API| Backend
Loading

Docker Images / 镜像下载(已经构建好)

  • AI Algorithm Image / AI 算法镜像: docker pull dingjianchen/gb28181-ai:latest
  • GB28181 Program Image / GB28181 程序镜像: docker pull dingjianchen/gb28181:latest

📸 Screenshots / 系统截图

1. Login / 登录

Login

2. Device Management / 设备管理

Device Management

3. Live Preview / 在线预览

Live Preview

4. Playback / 录像回看

Playback

5. Storage Settings / 存储设置

Storage Settings

6. Capability Center / 能力中心

Capability Center

7. Smart Alarm / 智能报警

Smart Alarm

✨ Features / 系统功能

1. Device Management / 设备管理

  • Auto-Registration: Automatically accepts GB28181 registration from IPCs and NVRs.
    • 自动注册：自动接受 IPC 和 NVR 的 GB28181 注册请求。
  • Channel Discovery: Automatically syncs device channels and status.
    • 通道发现：自动同步设备通道及在线状态。
  • Status Monitoring: Real-time Keepalive monitoring.
    • 状态监控：实时心跳保活检测。

2. Live Preview / 在线预览

  • Multi-Protocol Support: WebRTC (Low Latency), HTTP-FLV, HLS.
    • 多协议支持：WebRTC（低延迟）、HTTP-FLV、HLS。
  • H.265 & H.264: Supports modern H.265 (HEVC) codecs web playback (via Wasm or native support).
    • H.265 & H.264：支持 H.265 (HEVC) 编码的 Web 端播放（通过 Wasm 或原生支持）。
  • PTZ Control: (Supported via API) Pan, Tilt, Zoom control.
    • 云台控制：（API 支持）云台转动与变焦控制。

3. Cloud Storage / 云端存储

  • Server-Side Recording: Record streams to MP4 on the server.
    • 服务端录像：支持服务端 MP4 录像存储。
  • Storage Policy: Configure retention days and max storage usage.
    • 存储策略：配置录像保留天数及最大存储空间。
  • Auto-Cleanup: Automatically deletes old records to free up space.
    • 自动清理：自动清理过期录像释放空间。

4. AI Recognition / AI 智能识别

  • Object Detection: Built-in YOLOv8 integration for detecting objects (e.g., Person).
    • 目标检测：内置 YOLOv8 模型，支持目标（如：人）检测。
  • Smart Alarms: Generates alarms with snapshots when targets are detected.
    • 智能报警：检测到目标时自动生成报警并抓拍快照。

5. PTZ Control / 云台控制

  • Pan, Tilt, Zoom: Control camera movement and zoom levels via API.
    • 云台转动与变焦：通过 API 控制摄像头的移动和变焦。
  • Integration: Seamlessly integrates with the GB28181 protocol for real-time PTZ operations.
    • 集成：与 GB28181 协议无缝集成，实现实时云台操作。
  • Customizable: Supports custom PTZ presets and patrol routes.
    • 可定制：支持自定义云台预置位和巡航路径。

🚀 Getting Started / 快速开始

Prerequisites / 前置要求

  • Docker & Docker Compose

Installation / 安装部署

  1. Clone the repository / 克隆仓库

    git clone https://github.com/your-repo/gb28181-video-platform.git
cd gb28181-video-platform

  2. Configure Environment / 配置环境 Edit docker-compose.yaml to set your local IP address: 修改 docker-compose.yaml 设置本地 IP 地址：

    services:
  app:
    environment:
      - APP_GB28181_LOCAL_IP=192.168.1.100  # Your LAN IP
      - APP_GB28181_MEDIA_IP=192.168.1.100  # Your LAN IP

  3. Start Services / 启动服务

    docker-compose up -d --build

  4. Access the Platform / 访问平台

🤖 AI Service Usage / AI 服务使用

The AI service runs separately and connects to ZLMediaKit to process video streams for object detection. AI 服务独立运行，连接 ZLMediaKit 处理视频流并进行目标检测。

1. Prerequisites / 前置要求

  • Python 3.9+
  • CUDA (Optional, for GPU acceleration / 可选，用于 GPU 加速)

2. Startup / 启动服务

Method 1: Docker Compose (Recommended / 推荐)

Use Docker Image / 使用 Docker 镜像(已构建好)

Add the following service to your docker-compose.yaml: 在 docker-compose.yaml 中添加以下服务：

  ai-service:
    image: dingjianchen/gb28181-ai:latest
    container_name: gb28181-ai
    restart: always
    environment:
      - JAVA_API_HOST=http://192.168.1.100:8081 # Backend API address / 后端 API 地址
      - ZLM_HOST=http://192.168.1.100:8080      # ZLMediaKit HTTP API address / ZLMediaKit HTTP API 地址
      - RTSP_HOST=192.168.1.100                 # Local LAN IP / 本机局域网 IP
      - YOLO_MODEL=yolov8n.pt                   # Model: yolov8n.pt, yolov8s.pt, etc.
    volumes:
      - ./snapshots:/app/snapshots              # Map snapshots directory / 映射快照目录
Build and Run / 构建和运行（自己构建镜像）
# 构建镜像（注意这次会安装 ffmpeg 等库）
docker build -t gb28181-ai-service .

# 启动容器
docker run -id --name ai-service \
  -e ZLM_HOST=http://192.168.254.202:8080 \
  -e JAVA_API_HOST=http://192.168.254.202:8081 \
  -e RTSP_HOST=192.168.254.202 \
  -e RTSP_HOST_PORT=8554  \
  gb28181-ai-service

docker logs -f ai-service

Method 2: Source Code / 源码启动

cd ai-service

# Install dependencies / 安装依赖
pip install -r requirements.txt

# Run the service / 启动服务
# Ensure ZLMediaKit and Java Backend are running first!
# 请确保 ZLMediaKit 和 Java 后端已启动！
python app.py

3. Configuration (Environment Variables) / 配置（环境变量）

You can configure the service by setting environment variables before running: 启动前可通过环境变量进行配置：

Variable Description Default
JAVA_API_HOST Backend API Address http://127.0.0.1:8081
ZLM_HOST ZLMediaKit HTTP API Address http://127.0.0.1:8080
RTSP_HOST RTSP Stream Host (Local IP) 127.0.0.1
YOLO_MODEL YOLOv8 Model (n/s/m/l/x) yolov8n.pt
CONFIDENCE_THRESHOLD Detection Confidence (0.0-1.0) 0.5

⚙️ Configuration / 配置说明

Environment Variables / 环境变量

Variable Description (EN) Description (CN) Default
APP_GB28181_LOCAL_IP SIP Signaling IP SIP 信令监听 IP 0.0.0.0
APP_GB28181_MEDIA_IP Media Server IP (Advertised) 媒体流 IP（广播给设备） 192.168.x.x
APP_ZLM_BASE_URL ZLMediaKit API URL ZLMediaKit API 地址 http://127.0.0.1:80
YOLO_MODEL AI Model Path/Name AI 模型路径/名称 yolov8n.pt

🛠 Technology Stack / 技术栈

  • Backend: Java 25, Spring Boot 4.x, Maven
  • Frontend: React 18, Vite, TypeScript, TailwindCSS, Shadcn/UI
  • Streaming: ZLMediaKit (C++)
  • AI: Python 3.9+, PyTorch, Ultralytics YOLO
  • Database: SQLite (Easy deployment)

📄 License

MIT License

About

gb28181 轻量级平台

Resources

Readme
Activity

Stars

2 stars

Watchers

0 watching

Forks

0 forks
Report repository

Releases

No releases published

Packages

 
 
 

Contributors

Languages