Bird’s Eye View – Real-Time Multi-Camera Stitching for Embedded Vision at the Edge
RidgeRun’s Bird’s Eye View (BEV) solution transforms multiple video streams into a cohesive top-down perspective, enabling real-time situational awareness for embedded systems. Ideal for automotive surround view, autonomous navigation, robotics, and security applications, this software stitches camera feeds into a seamless aerial visualization—right at the edge, without needing external servers.
Leveraging your hardware’s on-board accelerators (e.g., NVIDIA Jetson), BEV enables rendering that’s optimized for embedded Linux environments.
What It Does:
-
Creates a real-time top-down (aerial) scene using input from multiple synchronized cameras.
-
Executes the stitching and blending using GPU or hardware accelerators.
-
Uses RidgeRun’s Libpanorama and the GstBEV GStreamer plugin to ensure smooth integration into multimedia pipelines.
Key Features:
-
Hardware-Accelerated Processing: Supports platforms with GPU or dedicated IP blocks (e.g., Jetson).
-
Flexible Calibration Tools: RidgeRun provides an intuitive GUI and CLI for spatial calibration of the cameras, enabling high-fidelity stitching.
-
Optimized for Embedded Linux: Perfect for real-time use cases on constrained hardware with minimal overhead.
-
GStreamer Integration: Seamlessly fits into existing video pipelines with RidgeRun’s GstBEV plugin.
-
Configurable Scene Layouts: Adapt the output view to vehicle outlines, scene contours, or application-specific geometry.
Example Use Case:
A large delivery truck equipped with four wide-angle cameras uses RidgeRun’s BEV plugin to create a real-time 360° view around the vehicle. The top-down stitched image is displayed in the cab, helping the driver detect blind spot obstacles during tight maneuvers in urban settings. The same BEV feed is also recorded for incident analysis—all processed on a Jetson Orin NX, using GPU acceleration to maintain low latency.
Ideal For:
-
Automotive ADAS and surround-view monitoring
-
AGVs, drones, and mobile robotics
-
Smart surveillance and command centers
-
Industrial automation with spatial video requirements
What's Included:
-
RidgeRun’s Bird’s Eye View
-
Libpanorama-based reference apps
-
GStreamer BEV plugin and sample pipelines
-
Calibration tool suite and documentation (binary)
-
Embedded software support via RidgeRun's consulting model
Consulting + Support:
This product is backed by RidgeRun’s Embedded Software Consulting team, offering design customization, integration, and long-term support for your BEV-enabled systems.
Learn more about Birds Eye View at our developer wiki.
Boards Supported
- NVIDIA Jetson
- x86 with NVIDIA dGPU