fgfg

docs/data_organization.md

@@ -7,13 +7,13 @@ The input and output of anTraX are organized in an  "experiment directory" (a.k.
 The input files are a sequence of raw videos, of the same animal group, recorded sequentially and under the same conditions. anTraX can process any [file format and video encoding readable by ffmpeg](http://www.ffmpeg.org/general.html#Supported-File-Formats_002c-Codecs-or-Features), however, all videos are expected to have the same file format, codec, frame size and framerate. Moreover, all videos must have the same base name, followed by a file index suffix seperated with a `_` character. 
 Optionally, for convinience, the videos can be organized into subdirectories. This can be a useful way to seperate  some meaningful partition of the experiment such as periods of consecutive recording, change in experimental conditions or feeding events, or just to partition a very long experiment.  The subdirectories should indicate the indexes of the videos stored in them. For example, subdirectory named "1_24" will contain videos 1 to 24.
 
-![expdir structure](/images/expdir_structure.png "structure of the experimental directory")
+![expdir structure](images/expdir_structure.png "structure of the experimental directory")
 
 ### The frame data files
 
 Optionally, each video will be accompanied by a frame information file, with the same name of the video and with a `.dat` extension. This file should contain a header row with the variable names (e.g. a timestamp, sensor data, etc.) and a value row for each frame in the video. If the file contains a variable named `dt`, it will be interpreted as the precise inter-frame interval, and will be used for tracking instead of the video framerate parameter.
 
-![dat file example](/images/dat_file_example.png)
+![dat file example](images/dat_file_example.png)
 
 To see an example for an experimental directory, download one of the [example datasets](datasets.md)
 

docs/datasets.md

@@ -7,61 +7,61 @@ As described in the anTraX publication, the software was benchmarked using 9 dif
 
 A clonal raider ant (*O. biroi*) with 16 ants tracked for 24 hours. This videos in this example are taken with a simple webcam, using low resolution (10 pix/mm, representing around 5x5 pixels per tag).
 
-![J16](/images/J16.png)
+![J16](images/J16.png)
 
 ### Dataset A36
 
 A clonal raider ant (*O. biroi*) with 36 ants tracked for 24 hours. This videos in this example are taken with FLIR-Flea3 12MP camera at a relatively high resolution (25 pix/mm) , enabling to distinguish between more colors, and also to augment the basic tracking with pose tracking using DeepLabCut (see anTraX publication).
 
-![A36](/images/A36.png)
+![A36](images/A36.png)
 
 
 ### Dataset V25
 
 A clonal raider ant (*O. biroi*) with 25 ants tracked for 6 hours. This videos in this example are taken with a simple webcam, but using higher resolution than in dataset J16. Although the resolution is high, the image quality in this example is reduced by an acrylic cover placed between the ants and the camera. This dataset is an example for tracking an open boundry arena, where ants can leave and enter through a specific part of the boundry.
 
-![V25](/images/V25.png)
+![V25](images/V25.png)
 
 
 ### Dataset G6X16
 
 This dataset is an example for tracking multiple colonies within the same video (6 colonies of 16 *O. biroi* ants). In addition, the image quality in this exampke is reduced by the low contrast between the ants and the background, the petri dish covers, and ligh reflections from those covers.
 
-![G6X16](/images/G6X16.png)
+![G6X16](images/G6X16.png)
 
 
 ### Dataset T10
 
 A colony of 10 *Temnothorax nylanderi* ants recorded for 6 hours using a webcam. The ants are marked with 4 tags each. 
 
-![T10](/images/T10.png)
+![T10](images/T10.png)
 
 
 ### Dataset C12
 
 A colony of 12 *Camponotous fellah* ants recorded for 6 hours using a webcam. The ants are marked with 3 tags each. Although the ants are big in this examples, their high velocity compare to the frame rate poses a challange. In addition, two ants are marked with the same color combination, but the algorithm was able to individually track them by using slight appearance variations. 
 
-![C12](/images/C12a.png)
+![C12](images/C12a.png)
 
 
 ### Dataset C32
 
 A colony of 12 *Camponotous* ants (unidentifies species) recorded for 24 hours. The ants are marked with 3 tags each. One ant (a virgin queen) is unmarked and identified by the classifier by other appearance markers (especially the wings).
 
-![C32](/images/C32.png)
+![C32](images/C32.png)
 
 
 ### Dataset D7
 
 A group of 7 fruit flies marked by one tag each and recorded for 1 hour.
 
-![D7](/images/D7.png)
+![D7](images/D7.png)
 
 
 ### Dataset D16
 
 A group of 16 fruit flies marked by two tags each and recorded for 5 hours.
 
-![D16](/images/D16.png)
+![D16](images/D16.png)
 
 

docs/jaaba.md

@@ -57,7 +57,7 @@ If your experiment contains more than one configured session, anTraX will run on
 
 anTraX will create a directory called `jaaba` under the session directory. In that directory, a subdirectory for each movie in the experiment will be created (JAABA consider each movie a separate experiment, and will therefore call each of these subdirectories an experimental directory). In each of these directories, there will be a soft link to the movie, named `movie.mp4` (or `movie.avi` etc. if your video files extension is different), and a mat file called `trx.mat` containing the trajectories in JAABA-compatible format. A subdirectory called `perframe` will also be generated, and will hold the perframe data.
 
-![expdir structure](/images/jaaba_directory_structure.png "jaaba directory structure")
+![expdir structure](images/jaaba_directory_structure.png "jaaba directory structure")
 
 
 ### anTraX-specific perframe features