Accelerated Robust Feature (SURF) is a robust image recognition and description algorithm, which was first published at the European Computer Vision Conference (ECCV) in 2006. This algorithm can be used in computer vision tasks, such as object recognition and 3D reconstruction. Part of his inspiration comes from SIFT algorithm. The version of SURF standard is several times faster than SIFT, and its author claims that it is more robust than SIFT in different image transformations. SURF is based on approximate 2D discrete wavelet transform, which effectively uses the integral graph.

The algorithm was first published in ECCV by Herbert Bay in 2006, and was officially published in Computer Vision and Image Understanding in 2008. This paper has been cited more than 9000 times. Heisenberg matrix is the core of SURF algorithm. For the convenience of operation, it is assumed that the function f(x, y) and the Hessian matrix h consist of the second-order partial derivative of the function: