 | 1 | initial version |
In David G. Lowe's paper, the Sigma of Gaussian weighting function is one half the width of the descriptor window. If the descriptor window is 3232 pixels with 44 subregions, I think the Sigma should be 16 rather than 2. But in sift.cpp, the code is "float exp_scale = -1.f/(d * d * 0.5f)", which means the Sigma is 2. Is there anything wrong?
| | 2 | No.2 Revision |
In David G. Lowe's paper, the Sigma of Gaussian weighting function is one half the width of the descriptor window. If the descriptor window is 3232 32x32 pixels with 44 4x4 subregions, I think the Sigma should be 16 rather than 2. But in sift.cpp, the code is "float exp_scale = -1.f/(d * d * 0.5f)", which means the Sigma is 2. Is there anything wrong?
| | 3 | No.3 Revision |
In David G. Lowe's paper, the Sigma of Gaussian weighting function is one half the width of the descriptor window. If the descriptor window is 32x32 pixels with 4x4 subregions, I think the Sigma should be 16 8 rather than 2. But in sift.cpp, the code is "float exp_scale = -1.f/(d * d * 0.5f)", which means the Sigma is 2. Is there anything wrong?
