pyhesaff 2.2.0

BUILDING

There are wheels published on pypi using cibuildwheel. You can install via:

pip install pyhesaff

TO build from scratch you will need development libraries for OpenCV and cmake (via scikit-build) should be able to find them.

On ubuntu 24.04 you can:

sudo apt-get install libopencv-dev

To get the required dependencies.

IMPLEMENTATION

Implementation depends on OpenCV (2.3.1+). Although, the code is original, the affine iteration and normalization was derived from the code of Krystian Mikolajczyk.

The SIFT descriptor code was patented under a US Patent 6,711,293, which expired on March 7th 2019, so the license is no longer required for use.

OUTPUT

NOTE THIS IS NO LONGER THE CASE. WE MAY REINSTATE THIS.

The built binary rewrites output file: <input_image_name>.hesaff.sift

The output format is compatible with the binaries available from the page “Affine Covariant Features”. The geometry of an affine region is specified by: u,v,a,b,c in a(x-u)(x-u)+2b(x-u)(y-v)+c(y-v)(y-v)=1. The top left corner of the image is at (u,v)=(0,0). The geometry of an affine region is followed by N descriptor values (N = 128).

File format:

N
m
u1 v1 a1 b1 c1 d1(1) d1(2) d1(3) ... d1(N)
      :
      :
um vm am bm cm dm(1) dm(2) dm(3) ... dm(N)

PROPER USE

If you use this code, please refer to

Perdoch, M. and Chum, O. and Matas, J.: Efficient Representation of Local Geometry for Large Scale Object Retrieval. In proceedings of CVPR09. June 2009.

TBD: A reference to technical report describing the details and some retrieval results will be placed here.

NOTES

Requires opencv. On ubuntu you can: sudo apt-get install libopencv-dev. You can also build / use wheels.