Video-Process for V-log

installation

pip install numpy matplotlib opencv-python mediapipe

project-explain

mask face in each frame in your video,

Now, I write down some function for processing one picture

workflow can be

1. use ffmpeg to generate png of each frame

2. proces each pic by python code

3. process pics to video without sound

4. ffmpeg merge video and sound

5. v-log is ok

example

from videoprocess.process import process_a_pic_by_face_solution, process_a_pic_by_holistic, process_a_pic_by_blaze
import os

process_a_pic_by_blaze('./masked_2304.png','zxm')

process_a_pic_by_face_solution('./masked_2304.png','zxm2')

process_a_pic_by_holistic('./masked_2304.png','zxm3')

We do not provide picture with human face, one can test with one's own picture.