Real-Time Object Detection with YOLOv8n and RTSP on Khadas Edge2 & VIM3
Explore AI vision at the edge with Khadas SBCs using YOLOv8n for real-time object detection and RTSP streaming. This solution is ideal for AI camera projects, smart surveillance, and low-latency video analytics.
Compatible with both USB webcams and MIPI cameras, this setup provides flexible hardware integration. The high-performance inference of YOLOv8n runs smoothly on Edge2 and VIM3/3L, with video output streamable via RTSP for easy access on any network-connected device.
Supported Devices & Guides:
- Edge2 – Object Detection Demo with RTSP
- VIM3/3L – Object Detection Demo with RTSP
Demo Preview – Real-Time Detection in Action
Below is a snapshot of the object detection system in action. The camera captures real-time video, and YOLOv8n detects multiple objects with bounding boxes and class labels. The processed video is streamed over RTSP and can be viewed on any compatible client.
