compute_grad.m

来自「张量投票在matlab环境下的仿真。球型张量及棒型张量的仿真」· M 代码 · 共 22 行

function grad = compute_grad(M,options)

% compute_grad - compute the gradient of an image using central differences
%
% grad = compute_grad(M,options);
%
%   'options' is a structure:
%   - options.h is the sampling step size on both direction (default 1).
%   - options.h1 is the sampling step size on X direction (default 1).
%   - options.h2 is the sampling step size on Y direction (default 1).
%   - options.type is the kind of finite difference.
%       type==2 is fwd differences, ie.
%           y(i) = (x(i)-x(i-1))/h, with special
%           care at boundaries.
%       type==1 is forward differences bilinearly interpolated in the
%           middle of each pixel (be aware that you have a shift of 1/2 on X and Y for
%           the location of the gradient).
%       type==1 is backward differences bilinearly interpolated in the
%           middle of each pixel (be aware that you have a shift of -1/2 on X and Y for
%           the location of the gradient).
%
%   Copyright (c) 2004 Gabriel Peyr