Hand Gesture Remote using Computer Vision And Rapsberry Pi

Hand gesture based remote

1. Introduction

Hand gesture based remote is an embedded device which will convert hand gestures to infrared signals (IR). It can recognize certain no of hand gestures of user and do some operations according to them.

Example - If we integrate with TV, we can increase/decrease volume, change the TV channels and etc.

We can integrate this system to control any kind of device/actuators.

The overall system has 4 main components

1. Camera Unit
2. Video Processing Unit
3. IR signal Unit
4. Power Unit

Here is the overview of overall system

1.1 Camera Unit

This unit will capture the video stream of the environment and send it to the video processing unit. Simply this unit contains camera and illumination system to capture video stream. This unit implemened using raspberry pi3 camera front illumination system.

1.2 Video Processing Unit

This unit will be done the main operation of the system. The received video data will be processed at this unit which was implemented using raspberry pi 3 development board. The image processing was implemented using OpenCV and python. The received video will process and detect hand gestures. Then detected hand gestures will be converted to predefined GPIO signal.

Here is a video demontrantion of the computer vison part
https://youtu.be/U8X_tLIOGAc

1.3 IR signal Unit

This unit contains two components.

          1.3.1 IR signal Generator

                   According to GPIO signal, this component will generate infrared signals (IR)

          1.3.2 IR signal Receiver

                   This component will capture IR signals generated by the IR generator and control the target actuators.

                   Example -  If we want to control a fan, we have to integrate this component with the switching circuit of the fan

1.4 Power Unit

The power unit will help to run the whole system. This will power the raspberry pi and IR signal unit. We use 5v/2A power supply in this system.