BubbleID Module
class DataAnalysis
This class is for generating and analyzing data from a single pool boiling experiment or specific video from an experiment.
- __init__(…)
- Parameters:
imagesfolder: path to folder where .jpg images are saved
videopath: path to where .avi video is saved
savefolder: path to folder where all data will be saved
extension: unique identifier that will be attached to each file saved in the save folder
modelweightsloc: path to location where model weights are saved
device: (‘cpu’ or ‘gpu’)
- GenerateData(…)
- Parameter:
thres: Value between 0 and 1 that specifies the cut off of confidence of detected bubbles that are kept
- Output:
bb-Boiling-output-{extension}.txt: File containing the coordinates of the bounding boxes of the detected bubbles and the confidence.
vapor_{extension}.npy: Numpy file containing the count of pixels containing a bubble. This file can be used for approximating vapor faction.
bubble_size_bt-{extension}.npy: Numpy file containing the number of pixels each detected bubble consists of for each frame. This can be used to determine bubble count and size.
bubind_{extension}.npy: Numpy file containing the labeled index of each bubble
frames_{extension}.npy: Numpy file containing the frames each bubble is present in
- Plotvf()
- Output:
vaporfig_{extension}.npy: Plot of vapor fraction over the duration of frames along with the rolling sample average
- Plotbc()
- Output:
bcfig_{extension}.npy: Plot of the bubble count over the duration of frames and rolling sample average
- PlotInterfaceVelocity(…)
- Parameter:
bubble: index of the bubble to find the interface velocity.
- Output:
displays an image of the frame and bubble contour to check if this is the intended bubble
velocity_{extension}_{bubble}.png: Figure of the interface velocity along the bubble contour vs frames bubble is present in. This figure has been smoothed with a gaussian filter and has a range of -30 to 30. This function is best used with higher framerates (i.e. 3000 fps) and larger bubbles.