Wireless Sensor Activated Camera

In this project I was able to make a wireless sensor activated security camera with the use of a raspberry pi 2 micro controller.

The motivation behind this project was to make a real time embedded system application.

From the established application idea, I had to make use of an adequate and reliable computing concept, to provide the best application performance based on time.

This embedded system was tested under three different computing concepts:

Baseline Code
Multitasking
Multi threading

I used the following components to make this project:

HC-SR04 Ultrasonic Range Sensor
Raspberry Pi 2
5MP Camera Module for raspberry pi
1KΩ and 500Ω Resistors
Jumper Wires

I used the following software components to make this project:

Python 3
Raspbian (Linux)
Visual Studio Code

System Diagram

The application was coded in a n infinite while loop to get continuous distance readings from the ultrasonic range sensor.

Based on the distance of 2-40 cm the application is triggered to take a picture of the object in range using the raspberry pi camera.

The picture taken by the application is then saved and sent over e-mail to the user.

Since the application is dependent on continuous distance readings, taken every second, the system had to have a time out to prevent overwriting of pictures and email spamming.

Source Code

Baseline.py

Multitasking.py

Multithreading.py

Demo Outputs

Console Output displaying running code

Output of file sent via email (sent by micro controller)

Output of opened email

System Component Wiring

Results

After running various trials of the application as the baseline, multitasking, and multi threading. I concluded that my wireless camera application was more precise and reliable when running as the baseline, based on states. This meant that this application was not behaving very fast when using multiprocessing or semaphores.

Baseline Execution Time

Multitasking Execution Time

Multi threading Execution Time

Google Sites

Report abuse