Multi-path RTSP video real-time display method and system
By employing a multi-channel RTSP video real-time display system on an embedded platform, utilizing hardware decoding and zero-copy transmission technology, and combining OpenGL and QML rendering pipelines, the problems of high CPU usage and large rendering latency in existing multi-channel RTSP video displays are solved, achieving low-latency, high-efficiency, and stable multi-channel video display.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-01-23
- Publication Date
- 2026-04-28
AI Technical Summary
Existing multi-channel RTSP video real-time display systems on embedded platforms suffer from problems such as high CPU usage, large rendering latency, multi-level copying of video frames, difficulty in connecting hardware decoding and rendering frameworks, and instability in multi-channel rendering, making it difficult to meet the requirements of multi-channel high-definition, low-latency, low-power, and highly stable display.
The system employs a multi-channel RTSP video real-time display system, which includes a video receiving module, a hardware decoding module, a GPU rendering module, and a rendering adaptation module. It receives multiple RTSP video streams in parallel, performs hardware decoding using hardware decoders such as MPP video decoding accelerators, and achieves zero-copy transmission using DMA-BUF and EGLImage technologies. It also seamlessly integrates OpenGL rendering pipeline and QML rendering pipeline.
It achieves low-latency, high-efficiency rendering and stable display of multiple video streams, reduces CPU computing burden, improves rendering efficiency, ensures the order of video streams and system stability, and is suitable for multi-channel video display scenarios such as mining, industrial monitoring and security.
Smart Images

Figure FT_1 
Figure FT_2