Closed under-score closed 1 year ago
Thank you for your suggestion. I'm adding it to my TODO list. I'll leave this issue open for now.
Could that table also be produced offline eg from Python CV2 / ORB?
StaticLSHTables
is something I've invented myself - for Speedy and the GPU. OpenCV has FLANN-based feature matching methods, though the implementations differ.
StaticLSHTables
is fed with an array of keypoints. It only has an output port, which is meant to be connected to a LSHKNN
node.
We now have a LSH feature matching demo available.
thank you!
would be great if we could have also a LSH version of https://github.com/alemart/speedy-vision/blob/master/demos/feature-matching.html that shows how to combine keypoint batches into one LSH table and run a query. Could that table also be produced offline eg from Python CV2 / ORB? Thank you.