amroamroamro

From c68ef0b7e946e4e915f0972a3e5eb6bf6a425f59 Mon Sep 17 00:00:00 2001
From: Amro <[email protected]>
Date: Fri, 23 Feb 2018 19:20:03 +0200
Subject: [PATCH] fixes for MinGW

* fix mingw detection in cmake file. When test is true, it defines
  a macro which instructs DShow header not to use secure string
  replacements not found in mingw
* dont pass gcc flags to windres which doesnt understand them
  (another workaround is to set ENABLE_PRECOMPILED_HEADERS=OFF)

amroamroamro

From 90cbdbb555eef10f4418574bc179c5106eece453 Mon Sep 17 00:00:00 2001
From: Amro <[email protected]>
Date: Wed, 15 Nov 2017 19:13:45 +0200
Subject: [PATCH] fixes for MinGW

* fix generation of pkg-config (libs shouldnt have .dll suffix)
* fix mingw detection in cmake file. When test is true, it defines
  a macro which instructs DShow header not to use secure string
  replacements not found in mingw
* dont pass gcc flags to windres which doesnt understand them

amroamroamro

From 37163b5ad31d1a7dffa3d4fdf65ce9a357d12054 Mon Sep 17 00:00:00 2001
From: Amro <[email protected]>
Date: Fri, 18 Aug 2017 16:19:26 +0300
Subject: [PATCH] fixes for MinGW

- add a define before including DShow header,
  to not use secure string replacements not defined by mingw
- dont pass gcc flags to windres which doesnt understand them
  (another workaround is to set ENABLE_PRECOMPILED_HEADERS=OFF)

amroamroamro

XML configuration files for compiling MATLAB MEX-files using VS2017. Tested on Windows 10 64-bit with MATLAB R2016b and Visual Studio 2017 (Enterprise Edition), but it should also work with the Community Edition.

1. First copy the files to MATLABROOT\bin\win64\mexopts (this might trigger a UAC prompt).
2. Next run mex -setup and mex -setup C++ in MATLAB, and select VS2017 compilers.
3. Finally test the new settings with a sample MEX-file:

>> mex -v -largeArrayDims test.cpp

amroamroamro

See #299.

Files:

MyClass.m
private/
    MyClass_.cpp

Compile the MEX-file as usual mex -largeArrayDims MyClass_.cpp (preferrably with C++11).

amroamroamro

Compare watershed function of MATLAB and OpenCV (see this SO question).

amroamroamro

Example of using CUDA in a MATLAB MEX-function.

Tested on Windows 8.1 x64, MATLAB R2015a, CUDA 6.5, Visual Studio 2013.

Steps to compile and test:

C:\> call "c:\Program Files (x86)\Microsoft Visual Studio 12.0\VC\vcvarsall.bat" amd64
C:\> set CUDA_PATH=C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v6.5
C:\> set PATH=%PATH%;%CUDA_PATH%\bin;%CUDA_PATH%\lib;%CUDA_PATH%\lib64

amroamroamro

Comparison of different array searching implementation in MATLAB

http://stackoverflow.com/a/1913831/97160

Two benchmarks are included, one for searching a single value, another for searching multiple values (vectorized version).

You could also try different matrix sizes, as well as changing the search values (compare best case where all values are found in the array vs. worst case scenario when none of the values are found).

amroamroamro

Python version of the MATLAB code in this Stack Overflow post: https://stackoverflow.com/a/18648210/97160

The example shows how to determine the best-fit plane/surface (1st or higher order polynomial) over a set of three-dimensional points.

Implemented in Python + NumPy + SciPy + matplotlib.

amroamroamro

Epipolar Geometry example

MATLAB code to show epipolar geometry with OpenCV (using mexopencv), based on code from OpenCV-Python Tutorials.