Skip to content

HTTPS clone URL

Subversion checkout URL

You can clone with HTTPS or Subversion.

Download ZIP
Official source code for TLD
Matlab C++ C

This branch is 14 commits ahead, 1 commit behind bilderbuchi:master

Improve the object detection feature

Now OpenTLD starts tracking when the object has a distance of 2 pixels
towards the border of the frame. In other words: it should now track
the WHOLE object in stead of a part of it.
latest commit 3fd502a8b3
@TheMrtN authored
Failed to load latest commit information.
_input first commit
_input_entrance Adding a dataset that shows the object detection feature.
bbox
img Correction on dimentions of grabbed frame in Octave
mex Compile with Octave on Linux (Ubuntu)
other Speed optimization in Matlab/octave discrimination.
test first commit
tld Add a new feature: automatically detecting an object using background…
utils Improve the object detection feature
.gitignore Added a feature that saves the object pictures to disk.
LICENSE first commit
README Add info regarding stream_server to README, switch off debug.
compile.m Speed optimization in Matlab/octave discrimination.
initcamera.m Correct some speed issues when using function exist().
run_TLD.m Adding a dataset that shows the object detection feature.
stream_server.c TLD can now grab webcam frames using Matlab without Image Acquisition…

README

################################################################################
# ATTENTION:
# This branch of the source code is a development version including Octave
# (http://www.gnu.org/software/octave/) compatibility measures, and is not 
# an official version!
# Tracking is  working in Octave now, with some changes w.r.t. Zdenek's branch
# See https://github.com/bilderbuchi/OpenTLD/issues/2 for more information.
#
#
# Webcam support: 
# This is working in Matlab WITHOUT Image Acquisition Toolbox
# through stream_server.c using sockets and OpenCV.
# You have to compile stream_server on your own for now, the command may look
# similar to this: gcc stream_server.c -lcv -lhighgui -o stream_server
#
# Webcam in octave is near completion. Frames are grabbed properly from the
# server but problems with mouse input arise when selectin object boudaries.
# For stream_server.c to work in octave, the package sockets should be present.
################################################################################

TLD (aka Predator)
----------------------------------------------------------------------------

TLD is an algorithm for tracking of unknown objects in unconstrained video streams. The object of interest is defined by a bounding box in a single frame. TLD simultaneously tracks the object, learns its appearance and detects it whenever it appears in the video.

1. License

This source code is released under the GPL license version 3.0.  For alternative licensing options contact the main author:  zdenek.kalal@gmail.com.

2. Project website

You can find more information about TLD at: http://info.ee.surrey.ac.uk/Personal/Z.Kalal/tld.html. This includes the description of TLD, links to research papers, posters and the licensing options.

3. Wiki

Many questions regarding TLD are already answered at the following wiki: https://github.com/zk00006/OpenTLD/wiki. These questions include installation and common errors. Make sure to check the wiki first.

4. Discussion group

If you do not find your answer in the wiki, ask the question directly at the following discussion group:  http://groups.google.com/group/opentld. There are currently around 250 participants and it is likely you will get the answer soon.  

5. Feedback

Predator learns from its errors; let us do the same in this community! Therefore, if you get an answer that was not covered in the wiki, please update the wiki so that other people do not have to face the same problem. The wiki is freely editable at the moment. 

6. Citations

In case you use TLD in an academic work, please cite the following paper:

@article{Kalal2010,
 author = {Kalal, Z and Matas, J and Mikolajczyk, K},
 journal = {Conference on Computer Vision and Pattern Recognition},
 title = {{P-N Learning: Bootstrapping Binary Classifiers by Structural Constraints}},
 year = {2010}
} 

             
Installation 
============

1. Windows
	Tested on: Matlab R2010a, VS2010, Windows 7 (32-bit, 64-bit), OpenCV2.2
	1.1 Install OpenCV
	    a) make sure to compile opencv 
	    b) make sure that system variable PATH includes the directory to the OpenCV DLLs 
	1.2 Check paths in 'compile.m' file
	1.3 run 'compile.m', if you get errors go to OpenTLD discussion group
	1.4 run 'run_TLD.m', TLD should track a motorbike
 
2. Mac
   ...

3. Linux
	1.1 Install OpenCV
		sudo apt-get install libcv-dev (this will most probably be 2.1)
	1.2a Install Matlab (>= R2009b). Possibly you will need some extension toolboxes
            include:
                Image Acquisition Toolbox
                Image Processing Toolbox
                Signal Processing Toolbox
                Statistics Toolbox
	1.2b There are efforts underway to make TLD work with Octave, an open source alternative to Matlab
		You will need to install octave >=3.4.
		Also, install octave extension packages miscellaneous, image, statistics from http://octave.sourceforge.net/packages.php
    1.3 Check paths in 'compile.m' file
    1.4 run 'compile.m', (either in matlab or octave). If you get errors go to OpenTLD discussion group
    1.5 run 'run_TLD.m', TLD should track a motorbike



----------------------------------------------------------------------------

(c) 2011 Zdenek Kalal, zdenek.kalal@gmail.com
Something went wrong with that request. Please try again.