Introduction to active vision

Sonar vision from first principles

For this lab you will make your own sonar vision system based on either a simple hardware-based lock-in amplifier, or, if you choose, a software-defined sonar.

If you choose the hardware option, you will use the system you previously made for lab 6, and extend it by building the world's smallest lock-in amplifier as its input stage. This will be used for sonar sensing devices such as the seeing aid for the blind. This system will also form the basis for the electric self-driving vehicle of future labs.

For Lab 7, we will make the following:

Wearable lock-in amplifier;
Wearable active vision (sonar) system;
Wearable vision system for the blind; or (optionally)
Wearable heart monitor.

If you choose the hardware option, simply follow the instructions; see the Instructable here: https://www.instructables.com/id/Miniature-Wearable-Lock-in-Amplifier-and-Sonar-Sys/

Don't worry if you can't finish all of it.

You can get full marks for finishing just the simple part of it!

SDSS = SDS2 = Software-Defined Sonar System

For the software equivalent you will implement a simple Doppler sonar system using a speaker and microphone. Generate a steady tone, e.g. 5,000 cylces per second, from a speaker. The same signal you send to the speaker will be mixed (multiplied) by what you recieve from the microphone. The result will be lowpass-filtered. The result will be displayed on a screen as a dot that moves up and down.

Please follow the outline presented in Tuesday's class.

Grading for software (SDSS) implementation:

Grading for hardware implementation:

Assembly of the lock-in amplifier prototype on breadboard, neatly wired, and well-done, 4/10

Wearable sonar system using the lock-in amplifier, 4/10

Capture and processing in Arduio of the output of the lock-in amplifier, 2/10

Bonus marks (for a grade higher than 10/10): Possible ideas:

Complex-valued lock-in amplifier;
X-Y plot of the complex-valued lock-in amplifier;
Demonstration of computer vision using machine-learning, e.g. as per Chapter 2 of the course texbook;
Computation of the ACT (Adaptive Chirplet Transform) and applying it to the output of the sonar;
Demonstration of ultrasound heart monitor using machine learning.