Skip to content

Commit d0f9d25

Browse files
committed
[docs] global: update documentation
1 parent 0410955 commit d0f9d25

File tree

4 files changed

+13
-0
lines changed

4 files changed

+13
-0
lines changed

docs/user/assets/background.png

221 KB
Loading
88.4 KB
Loading

docs/user/assets/multi_tracked.png

260 KB
Loading

docs/user/detection.md

Lines changed: 13 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,13 @@
1+
---
2+
id: detection
3+
title: Detection tips
4+
sidebar_label: Detection
5+
---
6+
7+
## Masking some regions
8+
9+
It is sometimes necessary to mask specific regions to avoid reflection or artifacts and enhanced detection. We will show a straightforward manner with the movie [ZFL_004]( http://data.ljp.upmc.fr/datasets/TD2/#ZFL_004) from the two-dimensional dataset.
10+
11+
1. Compute the background as usual. In this case, with will take the average. ![background](assets/background.png)
12+
3. Construct a mask using [Gimp](https://www.gimp.org/) or [Inkscape](https://inkscape.org/) by taking the background image and masking irrelevant areas (in black or white, depending on the system). ![background](assets/background_masked.png)
13+
5. Load the new background image in FastTrack and performs the tracking. ![background](assets/multi_tracked.png)

0 commit comments

Comments
 (0)