OpenTLD Installation for Matalb 2012a (64 bit)

Tested on: Matlab 2012a, Microsoft VS 2010, Win 7 (64 bit), OpenCV 2.4.2

See below.

(1) Install and Compile OpenCV 2.4.2 , then set PATH variables to link to OpenCV DLLs. (how to? see here)

(2) Setup mex compiler in Matlab. Run “mex –setup”, and select VS 2010 compiler form the list.

(3) Check paths in “compile.m”, edit it. Mine looks like so:

include = ' -IE:\OpenCV\install\include\ -IE:\OpenCV\install\include\opencv';
libpath = 'E:\OpenCV\install\lib\';
files = dir(fullfile(libpath,'*.lib'));

(4) In TLD source, comment out the following lines in lk.cpp, fern.cpp and bb_overlap.cpp. Otherwise  you will get an error.

#ifdef _CHAR16_T
#define CHAR16_T
#endif

(5) Run compile.m. If you get include errors, check PATH variables, and make sure “E:\OpenCV\install\bin;” included.

(6) Run run_TLD.m. If you get a 32-bit error, make sure your path ponits to the 64-bit version of OpenCV DLLs.

An example of errors--“lk.obj : error LNK2019: unresolved external symbol…”

BTW1, if you successfully compiled the mex files, but get a error like this:

"??? Unexpected Standard exception from MEX file……error: (-27) create() called for the missing output array in function create”

Error in ==>; tldTracking at 30
xFJ = lk(2,tld.img{I}.input,tld.img{J}.input,xFI,xFI); % track all points by
Lucas-Kanade tracker from frame I to frame J, estimate Forward-Backward error,
and NCC for each point

Error in ==>; tldProcessFrame at 25
[tBB tConf tValid tld] = tldTracking(tld,tld.bb(:,I-1),I-1,I); % frame-to-frame
tracking (MedianFlow)

Error in ==>; tldExample at 41
tld = tldProcessFrame(tld,i); % process frame i

Error in ==>; run_TLD at 43
[bb,conf] = tldExample(opt);

Please change the line 186 of lk,cpp from

cvCalcOpticalFlowPyrLK( IMG[J], IMG[I], PYR[J], PYR[I], points[1], points[2], nPts, cvSize(win_size,win_size), Level, 0     , 0, cvTermCriteria(CV_TERMCRIT_ITER|CV_TERMCRIT_EPS,20,0.03), CV_LKFLOW_INITIAL_GUESSES | CV_LKFLOW_PYR_A_READY | CV_LKFLOW_PYR_B_READY );

to

cvCalcOpticalFlowPyrLK( IMG[J], IMG[I], PYR[J], PYR[I], points[1], points[2], nPts, cvSize(win_size,win_size), Level, status     , 0, cvTermCriteria(CV_TERMCRIT_ITER|CV_TERMCRIT_EPS,20,0.03), CV_LKFLOW_INITIAL_GUESSES | CV_LKFLOW_PYR_A_READY | CV_LKFLOW_PYR_B_READY );

BTW2, Some Matlab extensions are needed:

Image Acquisition Toolbox, Image Processing Toolbox, Signal Processing Toolbox, Statistics Toolbox

For other platforms, pls refer to the link: https://github.com/zk00006/OpenTLD/wiki/Installation