README.md

trackdem (version 0.6)

cran version rstudio mirror downloads rstudio mirror downloads

Left: reconstructed trajectories of Daphnia magna individuals. Top right: identified fish (movie obtained from: DOI: 10.1109/CVPR.2015.7298992). Bottom right: the package contains functions to simulate moving particles.

--- The aim of **trackdem** is to obtain unbiased automated estimates of population densities and body size distributions, using video material or image sequences as input. It is meant to assist in evolutionary and ecological studies, which often rely on accurate estimates of population size, structure and/or individual behaviour. The main functionality of **trackdem** includes a set of functions to convert a short video into an image sequence, background detection, particle identification and linking, and the training of an artifical neural network for noise filtering. For more information on the package and usage, see our article that was published in Methods in Ecology and Evolution. --- ## Table of Contents * [Getting started](#getting-started) * [Installation](#installation) * [Tutorial](#tutorial) * [Updates](#updates) * [Version 0.5](#version-0.5) * [Version 0.4.3](#version-0.4.3) * [Version 0.4.1](#version-0.4.1) * [FFmpeg vs Libav](#ffmpeg-vs-libav) * [Bugs](#bugs) --- ## Getting started ### Installation **trackdem** can be installed from [CRAN](https://cran.r-project.org/web/packages/trackdem/index.html) or from github. wzxhzdk:0 To use the automated video to image and metadata creation function from **trackdem** users need Python>=2.7, FFmpeg (or Libav, as explained below) and ExifTool. Ubuntu users can paste the following commands in a terminal to install FFmpeg and ExifTool (Python should be included by default): wzxhzdk:1 Mac users can paste the following commands in a terminal to install FFmpeg: wzxhzdk:2 ExifTool can be downloaded from here. Follow the installation instructions for the OS X Package. The newest Python release, if not installed yet, can be downloaded here. Windows users can download Libav here (instead of FFmpeg, see below). Download the latest nightly-gpl release, and extract all files to a chosen location. Next, download the file named libgcc_s_sjlj-1.dll, and place it within the libav directory, in '/usr/bin'. ExifTool can be downloaded here. For ExifTool, download the stand-alone executable and place the exiftool(-k).exe file in a chosen directory. For convenience, you can change the name to exiftool.exe, as described in the installation instructions. Finally, Python can be downloaded here. Follow the instructions for installation. --- ### Tutorial A full tutorial is provided with the package (here). Some test code to get you started is given below. wzxhzdk:3

wzxhzdk:4

wzxhzdk:5

## Updates ### Version 0.6 * Add functions findPixelRange() and findMaxCost(). ### Version 0.5.2 * Add optional arguments 'logsizes' and 'costconstant' to trackParticles(). ### Version 0.5.1 * No longer requires R-package SDMTools ### Version 0.5 * Package is now compatible with Python 3 ### Version 0.4.3 * Create log file for function createImageSeq(). * Optimize functions to analyze greyscale images. ### Version 0.4.1 * Enable greyscale image sequences. * Optimized identification and tracking functions to work with longer image sequences. * Increased flexibility in function createImageSeq(). ## FFmpeg vs Libav Note that we previously used Libav instead of FFmpeg. However, Ubuntu no longer uses Libav, and users therefore need to install FFmpeg as described above. See e.g. here and here for more information on this confusing matter. Ubuntu/Mac users might need to provide the path to the FFmpeg executable, which can be obtained with the following terminal command: wzxhzdk:6 Use argument 'libavpath' in createImageSeq() and plot() to specify the correct path. On Windows, the provided instructions should still work. ## Bugs Bug reports and comments are ![welcome](https://github.com/marjoleinbruijning/trackdem/issues).

marjoleinbruijning/trackdem documentation built on Sept. 29, 2021, 7:52 a.m.