Insite Team v.2 Spring 2020
To see our updated project: CSCI-2345 Website!
Inside Team v.1 2019 Application Design:
Task: Create Development Plan, Requirements, Constraints
- Plan A
- Estimote Beacons
- Plan B
- Two general localization systems
- Sound
- Computer Vision
- One precision localization system
- NFC Tags
- Two general localization systems
Task: Create Sound System to Track Signals
- Four speakers in known locations around the room
- Each speaker produces a different frequency (19-22 kHz)
- iPhone picks up beeps about once per second from these speakers
- FFT to separate the beep frequencies
- Maximum likelihood point calculated
- X and Y grid points passed to app
Task: Test Signal Processing for Sound System
- Incoming audio is sampled at 44,100 times per second
- 1/400th seconds of audio is passed through FFT to create audio spectrum
- Audio magnitude data saved
- Signal processor looks for rapid increase in magnitude of key frequencies
- Arrival timestamps for each frequency are marked
- Arrival times used to calculate distances from speakers and most likely position of the phone
FFT: Fast Fourier Transform
Task: Create Computer Vision Tracking System
- Needed two cameras
- Camera calibration using chessboard
- Finding the parameters (extrinsic and intrinsic) of the transformation between an object in 3D space and the 2D image.
- People detection on each frame for each camera
- Triangulation from corresponding centroid pixel values to 3D world coordinates
- Send person location over TCP to the App Team
- Implementation
- Initially in Python with OpenCv
- Final working implementation all in Matlab
Task: Calibrate Tracking System in Gallery Space