# KCF tracker – parallel and PREM implementations
The goal of this project is modify KCF tracker for use in the
-[HERCULES](http://hercules2020.eu/) project, where it will run on
-NVIDIA TX2 board. To achieve the needed performance we try various
-ways of parallelization of the algorithm including execution on the
-GPU. The aim is also to modify the code according to the PRedictable
-Execution Model (PREM).
+[HERCULES][1] project, where it will run on NVIDIA TX2 board. To
+achieve the needed performance we try various ways of parallelization
+of the algorithm including execution on the GPU. The aim is also to
+modify the code according to the PRedictable Execution Model (PREM).
-Stable version of the tracker is available from [CTU
-server](http://rtime.felk.cvut.cz/gitweb/hercules2020/kcf.git.),
-development happens at [Github](https://github.com/Shanigen/kcf.).
+Stable version of the tracker is available from a [CTU server][2],
+development happens at GitHub [here][wsh] and [here][3].
+
+[1]: http://hercules2020.eu/
+[2]: http://rtime.felk.cvut.cz/gitweb/hercules2020/kcf.git
+[wsh]: https://github.com/wentasah/kcf
+[3]: https://github.com/Shanigen/kcf
## Prerequisites
-The code depends on OpenCV 2.4 (3.0+ for CUDA-based version) library
-and cmake is used for building. Depending on the version to be
-compiled you have to have [FFTW](http://www.fftw.org/),
-[CUDA](https://developer.nvidia.com/cuda-downloads) or
-[OpenMP](http://www.openmp.org/) installed.
+The code depends on OpenCV 2.4 library
+and [CMake][13] (optionally with [Ninja][8]) is used for building.
+Depending on the version to be compiled you need to have development
+packages for [FFTW][4], [CUDA][5] or [OpenMP][6] installed.
+
+On TX2, the following command should install what's needed:
+``` shellsession
+$ apt install cmake ninja-build libopencv-dev libfftw3-dev
+```
-SSE instructions were used in the original code and these are only
-supported on x86 architecture. Thanks to the
-[SSE2NEON](https://github.com/jratcliff63367/sse2neon) code, we now
-support both ARM and x86 architectures.
+[4]: http://www.fftw.org/
+[5]: https://developer.nvidia.com/cuda-downloads
+[6]: http://www.openmp.org/
+[13]: https://cmake.org/
## Compilation
This will create several `build-*` directories and compile different
versions in them. If prerequisites of some builds are missing, the
-`-k` option ensures that the errors are ignored. This uses
-[Ninja](https://ninja-build.org/) build system, which is useful when
-building naively on TX2, because builds with `ninja` are faster
-(better parallelized) than with `make`.
+`-k` option ensures that the errors are ignored. This uses [Ninja][8]
+build system, which is useful when building naively on TX2, because
+builds with `ninja` are faster (better parallelized) than with `make`.
-To build only a specific version run `make <version>`, for example:
+To build only a specific version run `make <version>`. For example,
+CUDA-based version can be compiled with:
``` shellsession
-make cufft
+$ make cufft
```
+[8]: https://ninja-build.org/
+
### Using cmake gui
```shellsession
$ git submodule update --init
$ mkdir build
$ cd build
-$ cmake [options] ..
+$ cmake [options] .. # see the tables below
+$ make
```
The `cmake` options below allow to select, which version to build.
|Option| Description |
| --- | --- |
| `-DASYNC=ON` | Use C++ `std::async` to run computations for different scales in parallel. This doesn't work with `BIG_BATCH` mode.|
-| `-DOPENMP=ON` | This option can only be used with CPU versions of the tracker. In normal mode it will run computations for differenct scales in parallel. In the case of the big batch mode it will parallelize the feature extraction and the search for maximal response for differenct scales. If Fftw version is used with big batch mode it will also parallelize Ffftw's plans.|
-| `-DBIG_BATCH=ON` | Concatenate matrices of different scales to one big matrix and perform all computations on this matrix. This mode doesn't work for OpenCV FFT.|
-| `-DCUDA_DEBUG=ON` | This mode adds CUDA error checking for all kernels and CUDA runtime libraries. Only works with cuFFT version.|
+| `-DBIG_BATCH=ON` | Concatenate matrices of different scales to one big matrix and perform all computations on this matrix. This mode doesn't work with `OpenCV` FFT.|
+| `-DOPENMP=ON` | Parallelize certain operation with OpenMP. This can only be used with `OpenCV` or `fftw` FFT implementations. By default it runs computations for differenct scales in parallel. With `-DBIG_BATCH=ON` it parallelizes the feature extraction and the search for maximal response for differenct scales. With `fftw`, Ffftw's plans will execute in parallel.|
+| `-DCUDA_DEBUG=ON` | Adds calls cudaDeviceSynchronize after every CUDA function and kernel call.|
+| `-DOpenCV_DIR=/opt/opencv-3.3/share/OpenCV` | Compile against a custom OpenCV version. |
-Finally call make:
-```
-$ make
-```
### Compilation for non-TX2 CUDA
to run it on different architecture, change the `--gpu-architecture
sm_62` NVCC flag in **/src/CMakeLists.txt** to your architecture of
NVIDIA GPU. To find what SM variation you architecture has look
-[here](http://arnon.dk/matching-sm-architectures-arch-and-gencode-for-various-nvidia-cards/).
+[here][9].
+
+[9]: http://arnon.dk/matching-sm-architectures-arch-and-gencode-for-various-nvidia-cards/
## Running
be `kcf_vot` binary.
It operates on an image sequence created according to [VOT 2014
-methodology](http://www.votchallenge.net/). You can find some image
-sequences in [vot2016
-datatset](http://www.votchallenge.net/vot2016/dataset.html).
+methodology][10]. You can find some image sequences in [vot2016
+datatset][11].
The binary can be run as follows:
bounding boxes of the tracked object in the format "top_left_x,
top_left_y, width, height".
+[10]: http://www.votchallenge.net/
+[11]: http://www.votchallenge.net/vot2016/dataset.html
+
### Options
| Options | Description |
| --visualize, -v[delay_ms] | Visualize the output, optionally with specified delay. If the delay is 0 the program will wait for a key press. |
| --output, -o <output.txt> | Specify name of output file. |
| --debug, -d | Generate debug output. |
-| --fit, -f[W[xH]] | Specifies the dimension to which the extracted patch should be scaled. It should be divisible by 4. No dimension is the same as `128x128`, a single dimension `W` will result in patch size of `W`×`W`. |
+| --fit, -f[W[xH]] | Specifies the dimension to which the extracted patch should be scaled. It should be divisible by 4. No dimension or zero rounds the dimensions to the nearest smaller power of 2, a single dimension `W` will result in patch size of `W`×`W`. |
## Authors
* Vít Karafiát, Michal Sojka
Original C++ implementation of KCF tracker was written by Tomas Vojir
-[here](https://github.com/vojirt/kcf/blob/master/README.md) and is
-reimplementation of algorithm presented in "High-Speed Tracking with
-Kernelized Correlation Filters" paper [1].
+[here][12] and is reimplementation of algorithm presented in
+"High-Speed Tracking with Kernelized Correlation Filters" paper \[1].
+
+[12]: https://github.com/vojirt/kcf/blob/master/README.md
## References
-[1] João F. Henriques, Rui Caseiro, Pedro Martins, Jorge Batista,
+\[1] João F. Henriques, Rui Caseiro, Pedro Martins, Jorge Batista,
“High-Speed Tracking with Kernelized Correlation Filters“, IEEE
Transactions on Pattern Analysis and Machine Intelligence, 2015
projects / hercules2020 / kcf.git / blobdiff