This paper presents a matching strategy to improve the discriminative power of histogram-
based keypoint descriptors by constraining the range of allowable dominant orientations
according to the context of the scene under observation. This can be done when the
descriptor uses a circular grid and quantized orientation steps, by computing or providing a
global reference orientation based on the feature matches. The proposed matching strategy
is compared with the standard approaches used with the SIFT and GLOH descriptors and …

1 Relation between vectors normalized with the L 1 and L 2 norms 6 2 Maximal L 1 distance
for L 2 normalized vectors......... 8 3 MROGH keypoint patch size test image............ 10 4 and
corresponding patches.................... 11 5 Non-planar dataset used for the evaluation........... 15
6 Images used to test descriptors on 2D rotations........ 16 7 Average number of correct
matches for 2D rotations...... 16