Motion-Detection-Based-on-KDE

Kernel Density Estimation, DA and ML homework in DHU.

This project is for simple motion detection based on KDE(Kernel Density Estimation).

kde.py

parameter mode

You can detect motion in a vedio or in a series of pictures by selecting parameter 'mode'.

parameter kernel

You can choose a kernel function between gaussian kernel function and epanechnikov kernel function.

test.py

usage

This file is used to tune threshold of final binary image quickly.

result.zip

content

This compressed file includes some of results after running kde.py .

md_data.zip

content

This compressed file includes some source material that can be used to motion detection.

source


grass: From lesson DA and ML in DHU given by Prof. Dawei Li.

ped: From Bilibili BV1fE411w7ac, uploader: @在下胖头陀 .

cr: From Bilibili BV1Mp411f7XW, uploader: @紅日浮茶 .