Skip to content

Latest commit

 

History

History
82 lines (47 loc) · 4.08 KB

README.md

File metadata and controls

82 lines (47 loc) · 4.08 KB
Raw

ratognize

ratognize is a command line video-processing tool for colored blob tracking, optimized for simple and fast execution on large scale video footage with many individual blobs to track. ratognize is based on OpenCV.

History

ratognize was developed at Eötvös University, Department of Biological Physics, throughout the EU ERC COLLMOT Research Grant for tracking painted animals (rats first, hence the name) for several hours, days or even months (don't worry, they were safe and treated well, we are still friends).

It got public on GitHub in line with open-access efforts after our publication of the following article (please cite it if you use this repo):

Synergistic benefits of group search in rats. (2020). Máté Nagy, Attila Horicsányi, Enikő Kubinyi, Iain D. Couzin, Gábor Vásárhelyi, Andrea Flack, Tamás Vicsek. Current Biology. DOI: https://doi.org/10.1016/j.cub.2020.08.079

Disclaimer: The code was created for researchers by researchers, sorry if it does not meet your coding standards. Customer service is available though 0-24h ;)

Install

Prerequisites

  • The code relies on OpenCV 4.x, you should install that first.

  • Using CUDA is also recommended for speedup.

  • Finally, ffmpeg is also needed to get proper video outputs.

The code base was tested under Linux (with gcc + cmake + make) and on Windows (with Visual Studio and with Visual Studio Code + cmake), and should run smoothly, although I know it never does. Feel free to ask and please help me if you have suggestions for improvement.

Linux

To use ffmpeg with CMake smoothly, you should first install the followings:

sudo apt install -y libavcodec-dev libavformat-dev libavdevice-dev libavfilter-dev

If you have installed all prerequisites, just run bootstrap.sh, go to the build directory and run make.

Windows

The code was first developed under Visual Studio 2017. There you need to setup your environment properly. The file called user_macros.props might be of help.

The build process also got tested recently in Visual Studio Code with cmake. For linking OpenCV properly, do not forget to set the system environment variable OPENCV_DIR to the proper value (run the following with your proper OpenCV path from a terminal: setx -m OPENCV_DIR c:\opencv\build\).

Usage

ratognize is a command line tool, that might open OpenCV windows during execution.

ratognize is the second element of a toolchain that should be used properly to perform multiple colored object tracking on multiple videos (possibly on supercomputers) simultaneously and smoothly. Here is the total recommended workflow:

  1. Create your color definitions with ColorWheelHSV
  2. Setup and run ratognize to get .blob data for all your defined colors.
  3. Run trajognize to get .barcode data for all the detected blobs.
  4. Run trajognize statistics to analyze your data automatically.
  5. If you have multiple video files to analyze, run trajognize sum to summarize data for all individual threads.
  6. Run trajognize plot to visualize data in several ways.

For minimal usage instructions of ratognize, run ratognize --help.

To setup ratognize properly, create a copy of the sample ini file, etc/configs/ratognize.ini, read it over thorougly and change parameters as needed for your project. After, use your own ini file with the --inifile argument when running ratognize.

For any further questions on usage please contact.

Definitions

  • blob - a blob is a single solid-colored area on an image that can be detected by ratognize with high efficiency.

  • barcode - a barcode consists of several blobs next to each other in a row. each object you want to track must have a colored barcode on it, consisting of 2-3-4-5 blobs, depending on the total number of objects to be tracked. Barcodes should be unique, even if read backwards (RGB == BGR).