npra / kattvhask Goto Github PK

Structure the output files into a folder structure organized as such:

2018:
  jan:
    01:
      video001.avi
      video002.avi
    02:
      video003.avi

This will allow users to visualize the camera angle and the regions of interest to be certain that the current setup is ok.

The webpage can be served through a standard nginx setup (apt-get install nginx) on the raspberry pi. Then the webpage (html5 + canvas + websocket). It needs to open a websocket to the local server and retrieve frames and the regions of interest.

I'm thinking of just painting it to a javascript canvas to make it easier.

Enable remote fetching of output videos in a secure fashion - and possibly also other control mechanisms.

Using a input configuration file to control the execution of "kattvhask".

Should give parameters as input device, output path, motion dection coordinates, etc.