About project

In this project a model is trained to know if a person is standing or sitting or walking which is based on 18 features calculated on 100 samples taken every 2 seconds.

features calculated on nodeMCU(configered with gy87 or gy521) is then send to pc through TCP/IP and there activity is recognized using a pretrained model.

Quick start

1)Install nodejs.
2)burn nodeMCUtoserver.ino code to your device enabled with wifi and gy87 or gy521 or any other suitable device.
3)Turn on hotspot of your PC and connect your nodeMCU to your hotspot(you need to edit the credential in the code).
4)copy IP from serial window and paste it in the tcp_ip.py
5)Now open new cmd window and run the command "python tcp_ip.py".
6)Now open cmd and run the command npm install(this might take some time).
7)Now run the command "npm start".

(once you have the IP of nodeMCU you can connect the device to just the power source)

For more information about the project refer to the branch v1_readme of this repository.

For any help feel free to mail me