Marangoni_flow_analysis

Data extraction and flow visualization for the marangoni flow project by Stefan Puttinger and Mahdi Saeedipour. The code was developed and released by Michael Weiß

Description

The current release could extract the diameters of a droplet as shown in the table below. The scripts also provides a video Handler for dealing with video and extraction of data:

Acquired frame	Processed frame	Abstract

One single frame can be described by the following parameters, which are explanied by the abstract of Stefan and Mahdi

• Center of droplet $c_x$ and $c_y$
• Diameters/Radius
  • Spreading diameter/radius $r_s$
  • Bursting diamters/radius $r_b$
  • Core diameter/radius $r_c$
• Fingerlength $\lambda$
• Droplet distribution $d_{sd}(r)$

For explanation see the abstract of Stefan and Mahdi

How to use the scripts?

Functions

The folder maragoniProcessing contains the following scripts:

• videoHandling.py
• dataExtraction.py

Import

To import the scripts the following line needs to be imported on top of your main file. To call a specific function write first scriptname followed by the function name.

from maragoniProcessing import *
video = videoHandling.videoHandler()

In the code snippet the module ist loaded and a videoHandler class from the videoHandling.py file is instantiated.

Video Handling

Calling the videoHandler class a filedialog is opened. By selection of the expirement the video variable is created

The variable has some specific functions:

# get the frame at Index=100
frame = video.getSingleImage(frameNumber=100)

# iterate over the frames
for frame in video:
    print(frame)

# transformation function -> Example OTSU
def otsu(frame):
    gray = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)
    th,otsu = cv2.threshold(gray,0,255,cv2.THRESH_BINARY_INV+cv2.THRESH_OTSU)
    img = cv2.cvtColor(otsu, cv2.COLOR_GRAY2RGB)
    return img

# creates an output video by a user defined function
video.creatOutputVideo(fcn=otsu, vidDim=None)    

The video.createOutputVideo can also take video dimension for cropping options. The function itself is a pipeline system shown in the flow chart

flowchart LR
    movie-->|frame|userDefiniedFunction\nExample:Otsu-->|frame|outputVideo

Loading

Data Extraction

Calibration

To get real world dimension a calibration image is needed. A good choice is a checkerboard with known grid and checker size. A good online generator is calib.io. Just import the function and you will get the scaling factor and a verification image

scal, calibImg = scalingFactor(
    calib_img=calibImg,
    patternSize=(4,4),
    checkerSize=10)

Radius

The function centerAndAllDiameters() from the dataExtraction.py script returns the parameters $c_y$, $c_y$, $r_c$, $r_b$, $r_s$ as a tuple. They can afterwards be extracted as a csv file

for img in video:
    print(dataExtraction.centerAndAllDiameters(frame=img))