1. Antarstick
   1. What is Antarstick
   2. Features
   3. Installation
   4. User guide

Antarstick is a program for analyzing height of snow in photos. The program processes time-lapse photos from a fixed camera capturing wooden stakes with a known length. In each photo the stakes are used as a reference to estimate the snow height at the place of each stake.

• simple graphical user interface
• automatic suggestion/detection of stakes
• stake editing
• option to link cameras that capture scene from different point of view potentially resulting in one stake being captured multiple times
• automatic stake aligning after stick and camera movements

1. Install Qt5

   Windows: download from https://www.qt.io/download-qt-installer

   Ubuntu-based OS: sudo apt-get install qt5-default

2. Install OpenCV Windows: download opencv-4.2.0-vc14_vc15.exe from https://github.com/opencv/opencv/releases/download/4.2.0/opencv-4.2.0-vc14_vc15.exe

   For Linux, please refer to the OpenCV installation guide at: https://docs.opencv.org/master/d7/d9f/tutorial_linux_install.html and replace occurrences of https://github.com/opencv/opencv/archive/master.zip with https://github.com/opencv/opencv/archive/4.2.0.zip

3. Install Python 3 https://www.python.org/downloads/release/python-391/

1. Clone this repo by running

   git clone https://github.com/mrazr/antarstick
   

   or download and extract the zip archive by clicking Code -> Download ZIP at the upper part of the page

2. Navigate terminal/command line to the root directory(antarstick) of the project

3. Install requirements

   1. On Linux it should suffice to run the script antarstick_install_linux.sh

   2. Windows:

      1. Create a new python virtual environment:

         python -m venv venv
         

      2. Install requirements:

         ./venv/Scripts/pip3.exe install -r requirements.txt
         

         If you use command line instead of powershell/windows 10 terminal then substitute forward slashes with backward slashes

1. Still inside the root directory, run the application by executing

   ./venv/bin/python3 mainwindow.py
   

   in terminal on Linux, or

   ./venv/Scripts/python.exe mainwindow.py
   

   in Powershell on Windows.

2. You’ll be presented with options to open a camera or a dataset. Press Add camera and choose a folder that contains photos.

3. On the left side of the application you’ll find a list of photo names and on the right side, occupying most of the application is the photo viewer. By clicking on the name of an image, the image will be displayed.

4. Find a photo with no snow, all sticks clearly visible, from top to bottom. The image should not be too bright nor too dim.

5. Press Find sticks in the top menu inside the photo viewer.

   1. Ideally, all stakes should be graphically marked with a green frame and circles with cross laying on top of each stake’s extremities.
   2. If some stakes are misaligned or you want to delete undesired stakes/stakes that are not really stakes, press Edit sticks in the top menu.
      1. Press the red x to delete a stake.
      2. Drag the circular handles to either move top/bottom endpoints or the middle to move the whole line.
      3. When you’re done exit the Edit mode by clicking the Edit sticks button.
   3. To ease the stick editing, you can zoom by scrolling and pan the view either by pressing the Middle mouse button and dragging or by simultaneously pressing Ctrl and Left mouse button and dragging.

6. Press Confirm sticks.

7. Press Process photos

   A menu will appear, where you’ll have an option to choose how many processes you want to spawn. Considering the analysis procedure is relatively computationally demanding, bear in mind that choosing a number of processes that matches the number of processes/cores of your CPU will have an effect on the responsiveness of the whole system.

   For now, due to the way Windows creates processes and handles their memory, the Windows version is limited to 2 processes per camera to avoid huge consumption of RAM.

8. The Measurement mode allows you to adjust incorrect measurements. Incorrect measurements can be corrected by clicking at the appropriate level in the sticks that have been assigned erroneous measurements.