Individual (yiwu2) -- Add GStreamer Jetson real-time video processing article

[yiwu2]

Summary

This PR adds a comprehensive guide on using GStreamer for real-time video processing on NVIDIA Jetson Orin platforms. The article focuses on hardware-accelerated video streaming for robotics applications, covering zero-copy video pipelines, network transmission, and integration with Python/C++ code.

Article Details

Title: Real-Time Video Processing on Jetson with GStreamer

Location: /wiki/networking/gstreamer-jetson-realtime-video.md

Category: Networking (focused on video streaming for teleoperation and remote monitoring)

Content Overview

The article provides:

1. Introduction to GStreamer on Jetson

   • Explains the performance benefits of zero-copy architecture using NVMM (NVIDIA Memory Map)
   • Details the three competing goals in robotics video: high-fidelity capture, low-latency perception, and efficient streaming
   • Performance metrics: 200-500ms → 10-30ms latency reduction, 80-90% → 5-10% CPU usage

2. Installation & Setup

   • Step-by-step installation instructions for JetPack 6.0-6.2
   • Verification steps and troubleshooting common issues
   • Plugin inspection and capability checking

3. Pipeline Examples

   • Camera to screen display pipeline
   • Zero-copy CUDA processing pipeline
   • Complete network transmission pipeline (UDP/RTP streaming)
   • Detailed explanations of each pipeline element and parameter

4. Code Integration

   • Python integration using PyGObject
   • C++ integration with CMake setup
   • Best practices and production considerations

5. Comprehensive References

   • Official GStreamer documentation
   • NVIDIA Jetson-specific resources
   • Tutorials, examples, and troubleshooting guides

Files Changed

• ✅ Created: wiki/networking/gstreamer-jetson-realtime-video.md (580 lines)
• ✅ Updated: wiki/networking/index.md (added article link)
• ✅ Updated: _data/navigation.yml (added navigation entry)