Image Stitching for NVIDIA Jetson

Image Stitching for NVIDIA Jetson – Real-Time Multi-Camera Stitching with Seamless Blending Image Stitching for NVIDIA Jetson is RidgeRun’s GPU-accelerated GStreamer plugin designed to combine multiple video inputs into a single panoramic or 360° view—with seamless blending at the borders for a polished, professional result. Built for high-performance embedded applications, this plugin transforms adjacent fisheye or rectilinear images into continuous visual streams using RidgeRun’s optimized Stitcher and Projector components. Ideal for robotics, 360° cameras, autonomous vehicles, and surveillance, the plugin supports real-time stitching at resolutions up to 4K at 30fps for three input sources, and integrates directly into any GStreamer pipeline. Key Features: Real-Time 360° or Panoramic StitchingMerge multiple fisheye or rectilinear video streams into one stitched output. Seamless Border BlendingAdvanced algorithms ensure clean transitions at stitch lines, eliminating visible seams. Supports Equirectangular ProjectionsOutput video in formats suitable for panoramic viewing or further processing. Built for GStreamer PipelinesIntegrates via the stitcher and projector elements—ideal for embedded multimedia pipelines. GPU-Accelerated on Jetson PlatformsDelivers high throughput and low latency using CUDA-based parallel processing. Example Use Case: An autonomous delivery robot uses four wide-angle cameras to navigate crowded urban environments. RidgeRun’s Image Stitching plugin is used to combine these camera feeds into a single, seamless 360° view, allowing onboard AI models to process the environment as a single frame—reducing complexity and improving situational awareness in real time. What’s Included: RidgeRun Image Stitcher and Projector plugins license GStreamer integration with example pipelines Tools for calibration and projection configuration (binary). Optional 11-hour Embedded Software Support package Supported Platforms: NVIDIA Jetson Nano, TX2, Xavier NX, AGX Orin x86-64 Linux with CUDA-enabled GPU (for development or testing) Why RidgeRun? With over 19 years of embedded software consulting experience, RidgeRun delivers production-ready multimedia frameworks optimized for real-time vision, streaming, and robotics applications. This plugin is part of RidgeRun’s larger ecosystem for camera processing and image transformation. Explore the developer wiki