Preprocessing

Code snippets from the PhD of Alexander Hunt. New, all intergrated version available at machine-code repository.

How to prepare the data for ML analysis

1. Process the data

To breakdown the videos into idividual frames, we use the following command from the root folder:

python -c "from code.convert import convertVideoToImages; convertVideoToImages('Data/Videos/', 'Data/Image_temp/')"

Then resize the frames to a fixed size.

python -c "from code.convert import resizeAllJpg; resizeAllJpg('Data/Image_temp/', (1080, 1920))"

Then random crop the images.

python -c "from code.convert import radomCrop; randomCrop('Data/Image_temp/', 'Data/Images/', (256,256))"

Now label the images cropped.

2. Split the data into training and test sets

3. Normalize the data

python -c "from code.convert import *; iterateBlur()"

python -c "from code.convert import *; detectBlurr()"

python -c "from code.convert import *; chopUpDataset()"