If inscribe
Learned some examples of feature point matching, found that these feature points are detected automatically, in addition to coordinate also contains Angle information, such as
But I need to match the two groups of feature points is marked by hand, just 2 d coordinates of points, so how to use the opencv in criticizing or some way?

In addition I marked feature points is similar to the braille point

CodePudding user response:

Oh, this is similar to how braille some feature points matching compare?

CodePudding user response:

Feature points matching cv2. DrawMatchesKnn ()

CodePudding user response:

The ORB feature point detection algorithm

CodePudding user response:

And image to extract the feature points matching is different, you said the ORB or other matching feature points, after to extract feature points and the feature points, and then you can use to describe information match, I said is the essence of lattice match, no description, is some point set out alone, such as braille

CodePudding user response:

Want to match or need to be described son ah, only some words will put some into a descriptor to match

CodePudding user response:

The

oh roar reference 5 floor? Response:

want to match or need to be described son ah, only some words will put some into a descriptor to match again cough up

Such a transformation descriptor? Warrior give a guide or tutorial?
As far as I know descriptor inside there are many other things, such as feature point a small area of the chart below, just like scene, I was the feature points like braille, is not the point of the image,