Binocular Image Stitching Module with GPU Acceleration (C++ / CUDA / OpenCV)

We are seeking an experienced Computer Vision / C++ Engineer to develop a binocular image stitching algorithm capable of fusing two fisheye camera feeds into a real-time panoramic image with minimal latency and distortion.

You will design and implement a binocular stitching module that includes:
Camera calibration and correction
Feature matching and image fusion
GPU-accelerated stitching (CUDA / OpenGL)
Unified C++ API interface for integration
JSON/yaml configuration for algorithm parameters
Cross-platform compatibility (Windows, Linux, ARM64)

Who We’re Looking For
Proven experience in image stitching / SLAM / panoramic vision
Expertise in GPU-based acceleration (CUDA kernels, memory optimization)
Ability to write clean, modular, well-documented code
Strong debugging and optimization skills
Bonus: experience with 5G, robotics, or remote visual systems

How to Apply
Please include:
A short summary of your experience in computer vision or GPU programming
Examples of similar projects (repositories, demos, or publications)
Estimated completion time and proposed budget
Optional: suggestions on algorithm optimization or architecture