Skip to content
No description, website, or topics provided.
C++ MATLAB
Branch: master
Clone or download
Latest commit 08dd62f Mar 13, 2019
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
MatchingMapMaker_tests @ b4f7aed
.gitmodules
README.md
compile.m
complexMulti.hpp
dff_glacier_production.m
mexInterrupt.hpp
mmm.cpp
test.m

README.md

MatchingMapMaker

The MatchingMapMaker (M3) is a MATLAB function used to detect shifts between two images using convolutions. The function will operate a weighted sum of absolute errors for each lag vector given as input.

The main code is written in C/C++, but the code is interfaced to MATLAB using mex.

The function mmm use 4 parameters:

  1. an odd size kernel, which can be full of one to compute the Sum of absolute error, or a normalized version to have the mean
  2. reference image
  3. compared image
  4. a matrix, where each line represents a potential lag vector.

and return for each pixel:

  1. the lag with the lowest error
  2. the lowest error

Usage in MATLAB:

[lagIndex,quality]=mmm(kernel,ref,shiftedImage,lagVector,'Algo');

Installation

  1. test with mex -setup c++ that the mingw-w64 compiler is installed, in other cases install it from here
  2. run compile
  3. run test (to test if it works! )
You can’t perform that action at this time.