18-549 Team 22
The goal of this project is to create a camera that can be used to visualize sounds in a three dimensional space. When designing new products companies look to find ways to reduce the amount of noise the product produces. By using our tool product designers can easily isolate the source of the noise to take steps to eliminate it.
The AcoustiCam is an imaging device that captures sound sources around an object then creates a 3-D modeled map showing the concentration of noise. It works by using microphones to collect sound data around the object and then applying signal processing techniques to create the heatmap. This heatmap of sounds data is overlaid on images of the object to find where the sound is coming from.
Acoustic Camera is company specializing in industrial-grade acoustic cameras. Their product line consists of microphone arrays ranging from 32 to 168 microphones for more precise measurements. Each type of array, either ring, dish, or square, may present a better use case for different situations. They also have over a dozen software plug-ins that allow for different visualizations of the collected signal.
The Acoustic Camera Nor848A by Norsonic is an all-in-one industrial grade acoustic camera. With up to 384 digital microphones, the Nor848A eliminates certain problems such as ghost-spots that result from a low number of microphones. By integrating the interface into the camera, the output can be easily attained through a direct LAN connection.
The near field acoustic camera by Microflown has been optimized to work in real operating conditions and in reverberant conditions. By using PU sound intensity probes, the camera only yields an accurate representation of the direct sound. In addition, the acoustic camera can be equipped with two types of analysis methods: direct sound field visualisation and near field acoustic holograph.
Since the main application of acoustic cameras is noise reduction, current cameras are used on vehicles and building structures to improve noise emission. However, most of these cameras are large and only capture the 2-D representation. Our project looks to take 3-D readings, which is much more precise and accurate, around smaller consumer products. Our motivation is to allow product designers and engineers to analyze these sound maps to create quieter appliances.
The AcoustiCam’s microphone array structure should minimize aliasing effects, partial reflections, and resonance effects.
Write scripts that will change the audio waveforms collected into a 3D map showcasing where and how much sound is present.
Develop a way to overlay the 3D model onto an image or video in order to show where sound is coming from in an intuitive way.
The primary use case of our product will be to profile sound in commercial objects. The acoustic camera will provide a sound profile of the object from different angles to help provide a 3D reconstruction of potential noise sources. Using a slider, a user will be able to select a range of frequencies of interest and see where they are located on a heatmap.
The interaction diagram describes the flow of how the whole system works from start to finish. To initiate the process, the user starts it by the GUI. The microphone and camera begin to capture data as the turntable rotates the target object. When the capture process is complete, the GUI updates with an image overlay that allows the user to interact with a slider to see the acoustic camera at different angles.
The TASCAM has eight Ultra-HDDA microphone inputs that can record at audio resolutions up to 96kHz/24-bit.
The Logitech C270 can capture up to 1280 x 720 pixels and will allow us to create an image overlay.
A simple servo that can rotate 360° with a torque of 30.6oz-in.
The microphone is a low power, high-performance device that consists of an acoustic sensor, a low noise input buffer, and an output amplifier.
The laptop will power the interface and the signal processing while allowing for an interactive GUI for viewing the acoustic camera results.
This PCB houses eight MEMS microphones, eight op-amps, and eight xlr microphone connectors. We custom made this PCB in order to use special MEMS microphones that would allow us to hear sound up to 80kHZ.
We used a continuous servo to rotate a platform that would hold an object, in this case, a phone. While rotating the platform, we would collect snippets of sound through our PCB. We would then send the sound to our TASCAM which we used as an ADC. From the TASCAM, it would go into Matlab where our script would conduct beamforming on the sound in order to replicate where the sound was coming from.
These two images represent how our Matlab script overlays a representation of the sound emitting from the speaker of the phone onto an image taken from our webcam.
