Just last week AMD released AMDOVX, their implementation of the Khrono Group’s OpenVX computer vision standard. Today NVidia has followed suit with VisionWorks. OpenVX is intended to enable developers to create computer vision application (think Kinect) that work on the devices GPU.
From the VisionWorks homepage:
NVIDIA VisionWorks toolkit is a software development package for computer vision (CV) and image processing. VisionWorks™ implements and extends the Khronos OpenVX standard, and it is optimized for CUDA-capable GPUs and SOCs enabling developers to realize CV applications on a scalable and flexible platform.
The core VisionWorks functions are engineered for solutions in:
- Robotics and Drones
- Autonomous Driving
- Intelligent Video Analytics
- Augmented Reality
— SNIP —
The toolkit helps you unlock the possibilities for GPU-based CV systems by adding your own algorithms and processing pipelines. Using VisionWorks in conjunction with other APIs like OpenCV provides access to many open-source CV algorithms.
- CUDA accelerated OpenVX 1.0.1 conformant API and NVIDIA extension primitives
- Framework for seamlessly adding user defined primitives
- Thread-safe API
- Example/sample pipeline code
- Documentation including Toolkit Reference Guide with Release Notes, Installation Guide, Tutorials and API Reference.
VisionWorks is available for Linux for Tegra, Ubuntu and Windows in Beta form.