BeagleVision will be a fast object recognition framework. The aim of the project is to have a hardware optimized implementation (utilising the DSP and NEON) of state-of-the-art local feature detectors and descriptors on the Beagle Board. BeagleVision will include our recent work on efficient feature detectors and descriptors, other well-known algorithms (e.g. FAST corner detector and SURF), and efficient matching algorithms for very large datasets (>1M images).
Comments are not currently available for this post.