CODEWITHSUNDEEP
opencvfeature-detectionbrute-force
Colin747
Colin747

Reputation: 5013

OpenCV BFMatcher - how does it classify?

I'm using the OpenCV ORB method to detect objects in a video stream. My question is to how exactly BFMatcher achieves this?

As far as I'm aware the FlannBasedMatcher uses a KD Tree to classify the objects and that this method will be faster than BFMatcher but the matches may not necessarily be the best.

Upvotes: 1

Views: 1595

Answers (1)

zedv
zedv

Reputation: 1469

The purpose of using BFMatcher is to match keypoint descriptors.

The simplest method of matching descriptors is searching for the most similar descriptor in different images. This is done by the BFMatcher::match method which finds the best match for each descriptor from a query image.

There are other methods such as BFMatcher::knnMatch and BFMatcher::radiusMatch that are more complex but must be used in certain applications.

BFMatcher OpenCV docs: http://docs.opencv.org/modules/features2d/doc/common_interfaces_of_descriptor_matchers.html#descriptormatcher-match

Upvotes: 1

Related Questions