This forum is disabled, please visit https://forum.opencv.org
 | 1 | initial version |
The descriptor has 32 or 64 or 128 values, depending on its type. So one descriptor is a vector of 32, 64 or 128 values. Storing it in a Mat, you'll have a Mat with one row and 32, 64 or 128 colums. The queryDescriptor object is a Mat that contains more descriptors (stored on rows), so every time you want to compute a match, that is computing the distance between the two descriptors. This is why you always do matching on rows.
And about the match method that returns a sorted list... I do not think so. In the docs they are saying not so.
